The management of comminuted or pathologic fractures in metaphyseal bone is a challenging problem for even very experienced fracture surgeons. Plates are typically positioned peripherally on the bone and as a result are subjected to eccentric loading. When the far cortex is deficient secondary to fracture, infection, or tumor, eccentric loading can lead to collapse through sequential screw loosening, resulting in hardware failure.
Fortunately, implants and techniques have been developed to help mitigate this problem. Fixed angle implants such as blade plates and dynamic condylar screws were used to provide rigid constructs that better withstand bending forces. As plate technology evolved, locking implants added an additional tool to the fracture surgeon's arsenal.1 These implants allowed technically easier application of fixed angle implants in periarticular fractures. However, in both settings, there is still significant eccentric load and the potential for varus collapse with hardware failure.2,3
In his seminal book, Dr Mast described a technique that has been used to help mitigate the problems seen by a missing far cortex. This principal-driven approach can be used to decrease and withstand eccentric loading and allow early functional weight bearing. He termed this “composite fixation” where an endosteal plate is used in combination with a standard cortical plate to substitute for the missing medial cortex.4 In this article, we present 10 cases of cortical substitution plating to effectively manage patients with large metaphyseal cortical defects.
This technique requires using a standard plate placed intraosseously to substitute for the absent far cortex. To accomplish this, 2 conditions must be met: the plate must be rigidly opposed to the intact portion of the far cortex and the intraosseous plate (IOP) and standard or extraosseous plate (EOP) must be rigidly interlocked. In this way, the 2 plates will create an “I-beam” that is ideally situated to resist bending and torsional forces with minimal biologic insult to the fracture.
The first step is appropriate preoperative planning to size the medullary canal to determine what size plate can be accepted into the canal. Once this is ensured, the fracture is exposed and the IOP is introduced through the defect or at the nonunion site. It is advanced in the direction of the midshaft (for a distal fracture, retrograde; for a proximal fracture, antegrade) until the IOP clears the fracture. The insertion direction is then reversed until the plate is impacted into the epiphyseal fragment while trying to maintain the plate as close to the endosteal portion of the far cortex. As in an I-beam, the distance between the IOP and EOP will determine its ultimate strength.
There are several possible techniques to deliver the IOP against the far cortex. A bone tamp or impactor can be very effective for this maneuver because it fits into plate holes nicely and the curve provides the correct angle to work through the defect and still effectively deliver the force from the mallet to the plate (Fig. 1). This force helps place the IOP both against the far cortex and distally into the epiphyseal segment. Once the plate is centered on the fracture, defect, or nonunion, then, the EOP is positioned and provisionally stabilized in a standard fashion. Additional compression of the IOP against the far cortex can be achieved once the EOP is positioned. A nonlocking screw can be placed that threads the near cortex and lands on the IOP in an area between screw holes so that it drives the plate against the endosteal portion of the far cortex. In some situations, a nut can be added to the screw and tightened against the IOP, broadening the surface area of contact. Alternatively, a locking screw can be placed in the EOP that is of appropriate length to push on the IOP as the locking threads engage. If this technique is used, it is very important to note that the screw should not just rest against the IOP but actually push it. It should be difficult to get the locking screw threads to engage in the EOP. If the defect is large enough and one has access, a small laminar spreader can be placed between both plates to maximize the tension and distance between both plates. Then nonlocking or locking screws can be placed as above to maintain the distance and tension.
The plates then need to be interlocked by placing screws from the EOP both proximal and distal to the fracture or defect that engage and thread through the IOP. These screws should be placed on an angle so that they barely fit through the screw hole and essentially jam the plate preventing torsion or shortening between both plates. Again, it is very important that the screws should not be placed straight across through both plates. Finally, a screw can be placed through the EOP at the diaphyseal end of the construct that abuts the IOP to prevent shortening. The remainder of the screws should be placed in the standard fashion as dictated by the fracture/nonunion that is being treated.
A 43-year-old woman was involved in a motor vehicle collision and sustained an open intraarticular distal femoral fracture with extensive metaphyseal and diaphyseal bone loss (Fig. 2). She was treated with aggressive debridement and irrigation until the wound was stable for definitive fixation. Her articular injury was reconstructed with small fragment screws, and the articular block was then connected to the remaining diaphyseal segment with a precontoured periarticular locking plate. The 14-cm bone defect was initially managed with a methylmethacrylate spacer in an effort to induce a Masquelet membrane (Fig. 3). At 6 weeks postoperatively, the cement spacer was removed and filled with autogenous bone grafts. Approximately 6 months after fixation, the construct failed with varus collapse. The patient was taken back to the operating room for fixation of her nonunion with a precontoured distal femoral locking plate. During this procedure, an IOP was added in an effort to mitigate the eccentric loads on the EOP (Fig. 4). The nonunion was further treated with placement of a combination of autogenous bone grafts from the contralateral femur (Reamer / Irrigator / Aspirator (RIA) intramedullary harvest; Depuy Synthes, West Chester, PA) and bone morphogenetic protein-2.
As per the technique above, the IOP was inserted with a bone tamp until it was seated medially and distally in the subchondral bone. A laminar spreader was then used between the EOP and IOP to tension the plate against the medial cortex. A locking screw was added to the EOP to maintain this tension (Fig. 4, mark A). A cortical screw was placed through the EOP abutting the proximal extent of the IOP (Fig. 4, mark B) before screws were added, both proximal and distal to the nonunion, that interdigitated between the EOP and IOP at oblique angles (Fig. 4, mark C). The patient was allowed to be weight bearing to tolerance at 6 weeks postoperatively. She returned for follow-up 1 year after her final surgery and is able to ambulate without assistive devices and minimal pain. Films taken at this visit show incorporation of her bone graft with remodeling of a neocortex medially (Fig. 5).
This case is an 82-year-old woman who fell from standing, sustaining a supracondylar/intercondylar fracture of her distal femur (Fig. 6). She was treated with anatomic reconstruction of her articular surface, followed by bridge plating of her distal femur using a locking periarticular plate (Fig. 7). She subsequently developed hardware failure with a varus nonunion 4 months after her surgery (Fig. 8). She was taken back to the operating room for composite fixation of her nonunion. A 10-hole 4.5-mm compression plate was contoured and placed in her defect using the technique described above. In this case, after impaction of the IOP with the bone tamp/impactor, the EOP was provisionally fixed and a locking screw was used to tension the plate against the medial cortex. Additional screws were added between the IOP and the stem of the total hip replacement to bolster fixation and buttress against proximal migration of the IOP (Fig. 9). The local bone graft was added with bone morphogenetic protein-2 to the fracture site. The patient returned for follow-up 2 years after treatment for her nonunion with a healed and remodeled fracture (Fig. 10).
Over the past 22 years, we have used this technique 10 times for complex fracture and nonunion reconstruction. This is obviously rare, and the indications were extensive bone loss, recalcitrant nonunion, and, most specifically, a fracture that could not be managed with alternative techniques. In this series, 7 of the fractures were in the distal femoral area, 1 was a tibial shaft nonunion, 1 was a fibular fracture in a pilon fracture, and 1 was an ulna fracture with an absent radial head. In all these scenarios, intramedullary implants were contraindicated because of intramedullary implants (total hip or total knee arthroplasties) or periarticular locations preventing intramedullary access.
The average age of the patients was 61.7 years (range 31.1–83.2 years). There were 4 men and 6 women. All the fractures healed after treatment with composite fixation. There were 2 cases that did require 1 additional surgery each. In 1 case, there was a superficial infection along the medial wound from a pilon fracture. The intraosseous technique in that case was used on the fibula and did not communicate with this wound. The patient did well after 1 debridement and is weight bearing without problems 5 months after surgery. The other infection required removal of the IOP from the proximal ulna with debridement and revision plating. This patient ultimately went on to heal well without signs of infection 7 months after her final surgery.
Composite fixation offers a technically demanding but fundamentally sound way to manage complex fractures and nonunions where the far cortex is either missing or incompetent to bear load sufficient for healing. Newer technologies do allow for fixed angle bridging constructs that are sufficient for healing in many scenarios. However, complex or recalcitrant nonunions, obese patients, and large defects all have high rates of failure with simple plating.
There are a number of solutions to the lack of medial-sided support. One would be simply another surgical approach with an additional plate placed. This is a straightforward and biomechanically sound approach. For large defects in the distal femur, the femoral artery may make this more difficult, although there are preliminary data that minimally invasive techniques may be safe.5 There is a small clinical series showing the efficacy of this technique if done before implant failure.6 It is unclear if that technique is safe in massive defects such as our first case. The biggest disadvantage of this technique is that it further compromises the periosteal blood supply in an area where early healing is desired. Another option is to use an autogenic or allograft fibular strut to replace the medial cortex.7,8 This is not as technically demanding as an IOP but is not without its challenges. There are some limited biomechanical data, however, that it may not improve construct strength compared with simple lateral locked plating.9
At our institution, we have used intramedullary nails as a means of cortical substitution with great success. It is technically less demanding than any of these other techniques but does have its limitations.10 For example, in the second case, an intramedullary implant was not an ideal option because of the femoral stem of a total hip replacement. It is also very difficult to get effective interlocking of the 2 implants, and therefore, the construct may be less stable. In another of our cases, there was a tibial shaft nonunion below a stemmed total knee replacement. More commonly though, comminuted intraarticular distal femoral fractures are a contraindication to the use of a retrograde intramedullary nail. Some periprosthetic fractures also may not be able to be treated with this construct because some total knee implants preclude the passage of a retrograde femoral nail. Therefore, this technique is limited to simple articular fractures, extraarticular injuries, some periprosthetic fractures, and nonunion surgery in which the articular block has healed. Composite fixation with an IOP is useful in all these scenarios.
There are some limitations to this report, notably the relatively small case series we have presented. These 10 cases we have performed all have progressed to healing, although 2 cases did have wound complications. The technique was used in a variety of anatomic locations, although distal femoral fractures or nonunions were the most common indication. The largest limitation is the technically demanding and time-intensive nature of the procedure. It is unclear how generalizable this technique is and what the “learning curve” would be for someone unfamiliar with composite fixation. Despite these limitations, composite fixation is a versatile and powerful technique to treat a variety of difficult fractures and nonunions where cortical bone loss has rendered traditional plating insufficient to maintain stability necessary for bony healing.
1. Kregor PJ, Stannard JA, Zlowodzki M, et al. Treatment of distal femur fractures using the less invasive stabilization system: surgical experience and early clinical results in 103 fractures. J Orthop Trauma. 2004;18:509–520.
2. Henderson CE, Kuhl LL, Fitzpatrick DC, et al. Locking plates for distal femur fractures: is there a problem with fracture healing? J Orthop Trauma. 2011;25(suppl 1):S8–S14.
3. Ricci WM, Streubel PN, Morshed S, et al. Risk factors for failure of locked plate fixation of distal femur fractures: an analysis of 335 cases. J Orthop Trauma. 2014;28:83–89.
4. Mast J, Jakob R, Ganz R. Planning and Reduction Technique in Fracture Surgery. Berlin, Heidelberg: Springer-Verlag; 1989.
5. Jiamton C, Apivatthakakul T. The safety and feasibility of minimally invasive plate osteosynthesis (MIPO) on the medial side of the femur: a cadaveric injection study. Injury. 2015;46:2170–2176.
6. Holzman MA, Hanus BD, Munz JW, et al. Addition of a medial locking plate to an in situ lateral locking plate results in healing of distal femoral nonunions. Clin Orthopaedics Relat Res. 2016;474:1498–1505.
7. Kumar A, Chambers I, Maistrelli G, et al. Management of periprosthetic fracture above total knee arthroplasty using intramedullary fibular allograft and plate fixation. J Arthroplasty. 2008;23:554–558.
8. Cole PA. Endosteal allograft plating for the treatment of recalcitrant nonunions. Tech Orthopaedics. 2003;18:344–355.
9. Makinen TJ, Dhotar HS, Fichman SG, et al. Periprosthetic supracondylar femoral fractures following knee arthroplasty: a biomechanical comparison of four methods of fixation. Int Orthop. 2015;39:1737–1742.
10. Attum B, Douleh D, Whiting PS, et al. Outcomes of distal femur nonunions treated with a combined Nail/plate construct and autogenous bone grafting. J Orthop Trauma. 2017;31:e301–e304.