An evidence-based description of aseptic tibial and femoral diaphyseal nonunions without segmental defects is based on a systematic search of MEDLINE. Aseptic nonunion of the femoral or tibial diaphysis without segmental defects and with an in situ nail, treated with reamed exchange nailing or augmentative plating and bone grafting, has consistently high union rates. Aseptic nonunion without segmental defects and with in situ plate and screw fixation is best managed with revision plate and screw fixation and autogenous bone graft. Various techniques and methods of biological stimulation have relatively high union rates.
Key points
- •
The presence of in situ hardware is a primary determinant of nonunion treatment.
- •
Aseptic diaphyseal nonunion of the tibia or femur without segmental defect and with an in situ nail are best managed with augmentative plating or exchange nailing.
- •
Aseptic diaphyseal nonunion of the tibia or femur without segmental defect and with in situ plate and screw fixation are best managed with revision plate and screw fixation and bone graft.
- •
Various bone graft methods, including intramedullary reaming, autogenous iliac crest bone marrow, and reamer-aspirator-irrigator technique, facilitate healing.
- •
Biologic implants (eg, recombinant human bone morphogenetic protein, platelet gel) and nonoperative treatments (ultrasound, electrical shock wave therapy) are associated with relatively high healing rates in most reports.
Introduction
This article describes an evidence-based approach to the treatment of aseptic tibial and femoral diaphyseal nonunions without segmental defects. The evidence for current best practices was obtained by searching the English-language literature for articles published from January 2005 through January 2015. The authors systematically searched Medical Literature Analysis and Retrieval System Online (MEDLINE) using the key terms: “fractures, ununited” or “nonunion” or “nonunions”, “tibia” or “tibial” or “femur” or “femoral”, and “treatment”. The search was limited to adults (age ≥18 years). The results were 860 potential articles. We then reviewed the abstracts of all of these articles to identify the relevant papers.
We excluded articles reporting treatment of only infected nonunions or segmental defects. We excluded case reports, review articles, and technique papers that did not report results of treatment. We excluded papers reporting treatment of failed arthroplasty, failed arthrodesis, tumor resection, nonunions involving the articular surface, metaphyseal or epiphyseal nonunions, nonunions following periprosthetic fracture, and nonunions following pathologic fractures. Finally, we excluded articles that had insufficient detail or contained large variation in the reported nonunion types or treatments. This process left 41 articles (25 femur, 22 tibia, 6 both) that we retrieved and reviewed in full.
Introduction
This article describes an evidence-based approach to the treatment of aseptic tibial and femoral diaphyseal nonunions without segmental defects. The evidence for current best practices was obtained by searching the English-language literature for articles published from January 2005 through January 2015. The authors systematically searched Medical Literature Analysis and Retrieval System Online (MEDLINE) using the key terms: “fractures, ununited” or “nonunion” or “nonunions”, “tibia” or “tibial” or “femur” or “femoral”, and “treatment”. The search was limited to adults (age ≥18 years). The results were 860 potential articles. We then reviewed the abstracts of all of these articles to identify the relevant papers.
We excluded articles reporting treatment of only infected nonunions or segmental defects. We excluded case reports, review articles, and technique papers that did not report results of treatment. We excluded papers reporting treatment of failed arthroplasty, failed arthrodesis, tumor resection, nonunions involving the articular surface, metaphyseal or epiphyseal nonunions, nonunions following periprosthetic fracture, and nonunions following pathologic fractures. Finally, we excluded articles that had insufficient detail or contained large variation in the reported nonunion types or treatments. This process left 41 articles (25 femur, 22 tibia, 6 both) that we retrieved and reviewed in full.
Femur
Augmentative Plating
Eight articles described the use of augmentative plating, the use of plate and screw fixation in addition to an in situ intramedullary nail, with autogenous bone grafting for aseptic femoral diaphyseal nonunions ( Table 1 ). All 147 cases reported in these 8 papers healed (100% union rate) with average times to union ranging from 4.3 to 7.5 months.
| Article | Number | Treatment Details | Union Rate (%) | Time to Union (mo) |
|---|---|---|---|---|
| Augmentative Plating | ||||
| Choi & Kim, 2005 | 15 | Augmentative plating with autograft | 100 | 7.2 |
| Birjandinejad et al, 2009 | 25 | Augmentative plating with autograft if <50% cortical contact | 100 | 4.8 |
| Chen et al, 2010 | 50 | Debridement, decortication, augmentative plating with autograft | 100 | 6.0 |
| Park et al, 2010 | 11 | Debridement, augmentative plating with autograft | 100 | 7.3 |
| Gao et al, 2011 | 13 | Debridement, augmentative plating with screws around nail, and autograft | 100 | 7.5 |
| Hakeos et al, 2011 | 7 | Augmentative plating with autograft; remove interlocking screws at 1 end of the nail, apply compression via the plate or tensioning device, relock the nail | 100 | 5.0 |
| Said et al, 2011 | 14 | Augmentative plating with compression and autograft in 9 cases | 100 | 4.3 |
| Lin et al, 2012 | 22 | Decortication, augmentative plating with autograft | 100 | 5.5 |
| Blade Plate Fixation | ||||
| de Vries et al, 2006 | 33 a | Hardware removal, blade plate with autograft or DBM in select cases | 97 | 5.0 |
| Intramedullary Nail | ||||
| Niedzwiedzki et al, 2007 | 22 | Hardware removal (if present), statically or dynamically locked intramedullary nailing, with reaming 0.5 mm larger than the nail | 59 | NR |
| Wu, 2009 | 18 | Hardware removal, debridement, 12 mm retrograde nail with 1 mm over-reaming, autograft | 89 | 4.2 |
| Megas et al, 2009 | 30 | Hardware (plate) removal, debridement, canalization, antegrade nailing with 1.5 mm over-reaming and static locking for hypertrophic cases or dynamic locking for atopic cases | 97 | 7.9 |
| Exchange Nailing | ||||
| Wu, 2007 | 74 | Exchange nailing, over-reaming ≥1 mm, statically or dynamically locked | 92 | 4.4 |
| Oh et al, 2008 | 11 | Statically locked exchange nailing (series included 1 infected nonunion) | 91 | NR |
| Shroeder et al, 2009 | 35 | Reamed exchange nailing with over-reaming ≥1 mm, new nail ≥2 mm larger than the original nail; statically or dynamically locked | 86 | 4.0 |
| Gao et al, 2009 | 5 | Reamed exchange nailing with nail 1–2 mm larger than the in situ nail | 100 | 7.8 |
| Naeem-ur-Razaq et al, 2010 | 43 | Statically locked exchange nailing | 91 | 5.0 |
| Park et al, 2010 | 7 | Reamed exchange nailing | 29 | 7.0 |
| Yang et al, 2012 | 41 | Reamed exchange nailing, new nail 1–3 mm larger | 78 | 6.8 |
| Swanson et al, 2015 | 50 | Exchange nailing, over-reaming ≥1 mm, statically locked, new nail ≥2 mm larger; subsequent dynamization required in 14 cases (28%) | 100 | 7.0 |
a 28 cases were aseptic and 5 cases were infected; the results of the infected cases were not reported separately.
In 2005, Choi and Kim reported a 100% union rate for 15 aseptic femoral nonunions treated with augmentative plating using an AO plate. In 2009, Birjandinejad and colleagues reported a 100% union rate for 25 aseptic femoral nonunions treated with augmentative plating using a lateral 4.5 mm broad dynamic compression plate (DCP) for midshaft nonunions with bone graft from the ipsilateral iliac crest for cases with less than 50% cortical contact (dynamic condylar screws or blade plates were used for proximal or distal nonunions but these were not reported separately from the overall series). The following year, Chen and colleagues reported a 100% union rate and good-to-excellent functional outcomes for 50 aseptic femoral shaft nonunions treated with augmentative plating using a broad 4.5 mm DCP. Also in 2010, Park and colleagues reported a 100% union rate for 11 aseptic femoral shaft nonunions treated with augmentative plating using compression plates.
Gao and colleagues reported a 100% union rate for 13 aseptic nonisthmic femoral nonunions treated with augmentative plating using locking plates. Hakeos and colleagues reported a 100% union rate for 7 nonunions treated with an augmentative plating technique that included removing the interlocking screws at 1 end of the in situ nail, applying compression intraoperatively via the plate or an articulated tensioning device, and then replacing the interlocking screws before DCP fixation. One subject had a postoperative infected hematoma and 1 had a residual leg length discrepancy. Said and colleagues reported a 100% union rate for 14 aseptic femoral nonunion treated with augmentative plating with a 4.5 mm broad DCP. In 2012, Lin and colleagues reported a 100% union rate for 22 femoral shaft nonunions treated with augmentative plating using a 4.5 mm broad DCP.
Blade Plate Fixation
In 2006, de Vries and colleagues reviewed a consecutive series of 33 aseptic subtrochanteric femoral nonunions that had a variety of prior failed methods of internal and external fixation. The subjects were treated with hardware removal, blade plate, and autologous bone graft (13 cases), or demineralized bone matrix (DBM; 10 cases). Five infected nonunions were included but were not reported separately from the overall series. Union was achieved in 32 of the 33 nonunions (97%).
Intramedullary Nail Fixation
We identified 3 articles reporting the use of intramedullary nailing to treat a total of 70 aseptic femoral shaft nonunions, with an overall union rate of 83% in 4 to 8 months.
In 2007, Niedzwiedzki and colleagues reported on 22 cases of aseptic femoral shaft nonunions treated with locked intramedullary nail fixation using nails 11 to 16 mm in diameter and 0.5 mm over-reaming. Although all cases had undergone 3 to 8 prior surgeries, these surgeries were not described except that 13 had failed at least 1 prior nailing. In addition, several cases were treated with either exchange nailing or with augmentative plating with intramedullary nailing; these cases were not reported separately. The union rate was only 59%. In 2009, Wu reported a union rate of 89% for 18 aseptic, atrophic supracondylar femoral nonunions with in situ plate and screw fixation treated with hardware removal, debridement, a 12 mm diameter retrograde nail with 1 mm over-reaming, dynamic locking, and autogenous bone graft. In 2009, Megas and colleagues reported a 97% union rate for 30 aseptic femoral shaft nonunions (25 atrophic) with an in situ plate treated with hardware removal, debridement, bone grafting in atrophic cases, and antegrade reamed intramedullary nailing with 1.5 mm of over-reaming. The nails were dynamically locked in 22 atrophic cases and statically locked in the 5 hypertrophic cases and in the 3 atrophic cases with 1 to 2 cm of shortening, for which the defect was filled with autogenous iliac crest bone graft at the time of intramedullary nailing.
Exchange Nailing
We located 8 publications reporting results of exchange nailing for a total of 266 aseptic femoral shaft nonunions, with an overall union rate of 89% with time to union ranging from 4 to 8 months.
In 2007, Wu reported a 92% union rate in 74 aseptic nonunions of the femoral diaphysis treated with exchange nailing including over-reaming by at least 1 mm. In 2008, Oh and colleagues reported successful union in 10 of 11 cases (91%) of femoral shaft nonunion (1 infected, result not reported separately) treated with reamed, locked exchange nailing with a nail diameter at least 1 mm larger than the removed nail. In 2009, Shroeder and colleagues reported a 91% union rate for 35 cases of aseptic femoral shaft nonunion treated with closed exchange nailing with over-reaming of at least 1 mm and new nail at least 2 mm (mean, 2.65 mm) larger than the initial nail. Gao and colleagues reported a 100% union rate in a mean of 8 months for 5 aseptic femoral shaft nonunions treated with exchange nailing.
In 2010, Naeem-ur-Razaq and colleagues reported on 43 aseptic femoral shaft nonunions treated with statically locked exchange nailing with a union rate of 91%. In 2010, Park and colleagues reported that only 2 of 7 (29%) aseptic femoral nonisthmal diaphyseal nonunions achieved union following exchange nailing with a nail at least 1 mm larger and over-reaming by at least 1 mm. Yang and colleagues reported union in 32 of 41 (78%) cases of aseptic femoral nonunion treated with exchange nailing with a 1 to 3 mm increase in nail diameter. Union was achieved in 27 of 31 (87%) isthmal cases and only 5 of 10 (50%) nonisthmal cases. In 2015, Swanson and colleagues reported a 100% union rate in a consecutive series of 50 aseptic femoral shaft nonunions (all with >50% cortical contact) treated with exchange nailing, including a statically locked nail with a diameter at least 2 mm larger than the removed nail and over-reaming of 1 mm. Nail dynamization was performed in 14 cases (28%) before attaining union.
High union rates can be attained with exchange nailing of aseptic nonunions of the femoral shaft, specifically for nonunions located in the isthmus. The technique includes over-reaming by equal to or greater than 1 mm, inserting a new nail with a diameter equal to or greater than 2 mm larger than the in situ nail, static interlocking using new screw locations to optimize screw purchase (eg, use a nail from a different manufacturer from the in situ nail), and nail dynamization in cases showing slow progression toward healing following exchange nailing.
Ilizarov Method
In 2010, Lammens and colleagues reported on 8 subjects with aseptic femoral shaft nonunion following reamed medullary nailing. The 4 cases with no segmental defect were treated with nail removal and Ilizarov monofocal compression. All subjects achieved union at an average of 7 months.
Other Methods
In 2013, Giannoudis and colleagues reported successful union in 13 of 14 cases (93%) of aseptic subtrochanteric femoral nonunions with an in situ broken intramedullary nail treated with nail removal; revision internal fixation (exchange nail or plate); and the diamond concept, which is combined use of recombinant human bone morphogenetic protein (rhBMP)-7, reamer-aspirator-irrigator (RIA) graft from the contralateral femur, and bone marrow aspirate.
In 2008, Calori and colleagues published a multicenter randomized trial that compared rhBMP-7 to platelet-rich plasma (PRP) for treatment of 113 aseptic long bone nonunions (including tibia, femur, humerus, and radius or ulna). Most subjects also received revision fixation. Results were not reported by bone; however, the overall union rate was 87% at an median of 8 months for the rhBMP-7 group and 68% at a median of 9 months in the PRP group.
In 2009, Cacchio and colleagues compared 2 intensities of extracorporeal shock-wave therapy (ESWT; 4 weekly sessions of 4000 impulses at either 0.4 mJ/mm 2 or 0.7 mJ/mm 2 ) with surgery (intramedullary nailing or external fixation) to treat long bone nonunions, including 34 femurs. Healing rates were very high (≥92%) in all 3 groups but were not reported by bone. Xu and colleagues reported a 64% healing rate among 22 atrophic or hypertrophic femoral nonunions treated with only a single session of ESWT (6000–10,000 impulses, 0.6 mJ/mm 2 ). In 2012, Roussignol and colleagues reported a healing rate of 92% for 12 femoral shaft nonunions treated by low-intensity (30 mW/cm 2 ) ultrasound for 20 minutes per day up to 6 months.
Summary for Femoral Nonunions
For aseptic femoral shaft nonunions without segmental defect and with an in situ intramedullary nail, augmentative plating with bone grafting was reported to have a 100% union rate in all 8 articles reviewed ( Fig. 1 ). Exchange nailing is also a good treatment option with slightly lower reported union rates ranging from 78% to 100%, with higher rates reported for nonunions in the isthmal region. For aseptic nonunions with an in situ plate, hardware removal followed by blade plate fixation or intramedullary nailing with autogenous bone graft has comparably high success rates (≥89% across 3 reports).
