Long-bone deformity correction is a four-step process that involves (1) performing the surgical approach, (2) osteotomizing the bone, (3) correcting the deformity, and (4) stabilizing the correction with a fixation construct. At each step, multiple options are available, yielding many procedural permutations. Over the years, the addition of new options has increased the complexity of planning for deformity correction1-13. In an effort to simplify long-bone deformity correction, we developed the clamshell osteotomy. This procedure consists of two transverse bone cuts with an intervening longitudinal cut. Functionally, the clamshell acts as a subtraction or bypass osteotomy by realigning the mechanical axis of the extremity. Technically, it takes advantage of the mechanical strength of a reamed, statically locked, intramedullary rod to maintain length, alignment, and rotation until healing has occurred. The purpose of the present study was to report the preliminary experience with the clamshell osteotomy at three institutions.
Materials and Methods
Inclusion and Exclusion Criteria
A protocol for the prospective evaluation of patients with diaphyseal deformities was begun in January 2003 and was approved by the institutional review boards at three centers. The inclusion criteria for the osteotomy were (1) a diaphyseal malunion of the femur or tibia, (2) patient dissatisfaction with both the aesthetic appearance and the functional performance of the limb, and (3) a patient description of progressive pain in adjacent joints14. The exclusion criteria were (1) metaphyseal deformities, (2) deformities not thought to be amenable to acute correction secondary to shortening of >5 cm, (3) a soft-tissue envelope preventing a surgical approach to the malunited segment, and (4) evidence of active infection.
Preoperative evaluation included a standard musculoskeletal examination, with lower extremity length discrepancies being measured with use of a tape measure placed from the anterior superior iliac spine to the tip of the medial malleolus. To further assist with length measurements, blocks of differing heights were used to level the pelvis in double-limb stance. When limb-length inequalities measured >2.5 cm, radiographic scanograms were made. Soft-tissue changes resulting from previous procedures or traumatic open wounds were noted, with specific attention being paid to the soft-tissue zone over the segment of the deformity. Range of motion and joint stability proximal and distal to the deformity were noted. The rotational profile was judged according to the supine resting posture of the extremity, a thigh-foot axis determination for tibial torsion, and the rotatory hip range of motion arc in the supine and prone positions for femoral deformity. Radiographic evaluation included characterization of the malunion with biplanar radiographs, an evaluation of joint status proximal and distal to the deformity with standard radiographs, and standing hip-to-ankle radiographs in the patella-forward position.
A summary of the deformity characteristics is presented in Table I. A brief history of each case along with a description of the deformity characteristics are provided in a table in the Appendix. Coronal and sagittal plane angulation was determined by drawing the angle formed by the intersection of the anatomic axis of the most distal segment with that of the most proximal segment of the bone involved. Rotation and translation were referenced according to the standard practice of describing the distal segment with respect to the proximal segment. The amount of translation in the coronal and sagittal planes was estimated on the basis of the perpendicular distance between the anatomic axis of the proximal segment and the anatomic axis of the distal segment when both a translational and angular deformity were present. Such deformity characterization provided an improved understanding of the preoperative situation. Although all characteristics were taken into account, the two most important questions were (1) Does the soft-tissue envelope provide a safe surgical passage to create the longitudinal osteotomy? and (2) Can an acute correction occur with limited risk of pathologically stretching the neurovascular structures?
Antegrade Femoral Rod Fixation
Three of the four patients with a femoral malunion were managed with antegrade femoral rod fixation. For this procedure, the patient was placed in the supine position on a fracture table with boot traction. A piriformis starting point and anatomic axis entrance angle were used. The proximal femoral segment was sequentially reamed to allow for easy manipulation of the ball-tipped guide rod. An incision was then made on the lateral part of the thigh, spanning the length of the malunited segment. The iliotibial band was identified and incised to reveal the vastus lateralis fascia. The vastus lateralis muscle was then elevated from the posterior part of the femur, extraperiosteally exposing the malunited segment. An atraumatic soft-tissue technique was emphasized. The proximal and distal extents of the malunion were verified with use of fluoroscopy. A 3.2-mm drill-bit was then used to drill multiple bicortical holes along the long axis of the malunited segment. The goal was to create a uniform line of stress-risers. A 1-in (2.54-cm) straight osteotome was used to osteotomize the segment through the near cortex only with use of the drill-holes as a guide. The femur was then transected perpendicular to the normal portions of the diaphysis just proximal and distal to the malunited segment with use of a sagittal saw, creating a free intercalary segment of bone. The free intercalary segment was wedged open initially with a 1-in (2.54-cm) osteotome and subsequently with a lamina spreader. The lateral cortex was separated widely to ensure that the medial cortex was also separated. This step of the procedure has been likened to opening a clamshell. Occasionally, separate fracture lines were accidentally propagated within the osteotomized segment during the opening of the clamshell. These were ignored, and the technique was continued.
After realignment of the proximal and distal femoral segments with use of indirect reduction techniques, a guide rod was placed from the center of the anatomic axis of the proximal segment through the osteotomized segment to the center of the distal femoral segment, with careful attention being paid to the entrance angle into the distal segment15. The vastus lateralis was then allowed to drape over the exposed segment in order to retain bone fragments produced by subsequent reaming. Medullary reaming was carried out in the proximal segment until the osteotomized segment was encountered. The reamer was then pushed through the osteotomy zone until the intact distal femoral fragment was encountered. Reaming of the distal fragment was then performed. A femoral rod measuring 1 mm less in diameter than the final reamer was then implanted. Interlocking bolts were placed in the proximal femoral fragment in the static position with use of the jig. Final length and rotational corrections were performed at this time on the basis of preoperative measurements, and then interlocking through the distal femoral segment with the freehand perfect circle technique was accomplished.
The vastus lateralis was then retracted to expose the osteotomy zone. The secondary gaps created by the restoration of length, alignment, and rotation were then inspected. The bone fragments produced by reaming generally filled the osteotomy gaps. In the cases of three patients, it was thought that the bone fragments produced by reaming may have been insufficient to adequately bridge the osteotomy gaps and demineralized bone matrix (DBX; Synthes, Paoli, Pennsylvania) was used to supplement the bone fragments produced by reaming. The iliotibial band and skin were then closed. A case example is presented in Figures 1-A through 1-F.
Retrograde Femoral Rod Fixation
One patient with a femoral malunion was managed with retrograde femoral rod fixation. For this procedure, the patient was placed in the supine position on a radiolucent operating table with both of the lower extremities included in the surgical field. The involved lower extremity was placed atop a radiolucent triangular positioning device, allowing the knee to be flexed approximately 45°. A medial parapatellar approach was used to gain access to the starting point. The starting point and the entrance angle into the distal segment were confirmed radiographically, followed by distal segment reaming. The malunited femoral segment was then treated exactly as described above with the antegrade rodding technique. The reaming rod was placed through the distal femoral segment, through the osteotomy segment, and up to the piriformis fossa in the proximal segment, with care being taken to ensure an appropriate entrance angle and ending point in the proximal segment15. Before sequential reaming, provisional alignment was achieved in all planes with use of indirect reduction maneuvers. After reaming, a retrograde femoral rod was placed and distal interlocking bolts were placed through the jig. Final length and rotational adjustments were performed with use of the contralateral lower extremity as a template, and then proximal interlocking bolts were placed.
Tibial Rod Fixation
The patient was positioned supine on the operating room table with both lower extremities in the surgical field. A lateral incision was created along the fibular shaft at the planned level of the proximal transverse component of the tibial osteotomy. An oblique osteotomy was created in the fibula to allow for a larger surface area for contact and healing. A medial parapatellar entrance to the previously defined safe zone for the tibial rod starting point was used16. With use of fluoroscopy, care was taken to ensure an appropriate entrance angle into the proximal tibial segment during reaming. A longitudinal incision was made over the anterior compartment, one fingerbreadth lateral to the tibial crest along the proposed longitudinal osteotomy site. The anterior compartment musculature was translated posteriorly, allowing for an extraperiosteal exposure of the lateral aspect of the malunited segment. An atraumatic soft-tissue technique was emphasized. The positions of the proximal and distal transverse osteotomies were localized with radiographic guidance. Unlike the coronal plane clamshell osteotomy that was performed in the femur, the clamshell component of the osteotomy was created parallel to the medial tibial face (Fig. 2). A 2.5-mm drill-bit was used to create the path for the longitudinal osteotomy with the goal of creating a bicortical uniform plane of stress-risers. Completion of the osteotomy of the near cortex only was accomplished with an osteotome with use of the drill-holes as a guide. A sagittal saw was used to create the transverse proximal and distal osteotomies. The far cortex of the osteotomized segment was split parallel to the medial face with use of an osteotome and a lamina spreader. The limb was placed over a radiolucent triangle. The guide rod was passed from the proximal segment through the osteotomized segment into the distal tibial segment under fluoroscopy. Care was taken to ensure that the entrance angle and the ending point in the distal segment were appropriate15. The anterior compartment was allowed to drape over the cortex to preserve the bone fragments produced by subsequent reaming at the osteotomy sites. The proximal and distal segments were reamed sequentially until cortical chatter was noted. The reamer was pushed through the clamshell segment to protect the neurovascular structures. The tibial rod was then passed, and standard proximal interlocking was accomplished. The leg was removed from the positioning pillow and was placed flat on the operating table. The other leg was used as a control for final length and rotational alignment, followed by distal interlocking. The anterior compartment musculature was retracted posteriorly from the lateral part of the tibia to allow for inspection of the osteotomy sites. The bone fragments produced by intramedullary reaming usually filled the osteotomy site. For gaps of ≥1 cm, demineralized bone matrix (DBX; Synthes) was used. Closure of the extensile approach to the malunited segment was achieved with use of the Allgower modification of the Donati technique17, emphasizing careful soft-tissue handling. A case example is presented in Figures 3-A through 3-F.
All patients were admitted to the hospital, where patient-controlled analgesia as well as oral narcotics were used for pain control. Intravenous cefazolin was administered for twenty-four hours postoperatively. Mobilization began on the first postoperative day, with the patient using crutches with toe-touch weight-bearing under the direction of a licensed physical therapist. Weight-bearing was advanced incrementally as osteotomy site healing progressed, with a goal of full weight-bearing by twelve weeks. Prophylaxis against thromboembolic disease was provided by the use of support stockings, sequential compression devices, and low-molecular-weight heparin. The patients were followed in an outpatient setting beginning two weeks postoperatively. Postoperative evaluation included clinical and radiographic evaluation at two weeks, six weeks, three months, six months, nine months, one year, and two years. Clinical evaluation included a standard extremity-based musculoskeletal examination with a focus on motion of the joints proximal and distal to the correction, the strength of the muscles around the previous deformity, gait pattern, and peripheral nerve status. The postoperative radiographs included biplanar views of the previously malunited bone. The radiographs were evaluated for progression of healing and maintenance of the anatomic axis alignment that had been achieved at the time of surgery. Biplanar radiographs of the joints proximal and distal to the correction and standing hip-to-ankle radiographs were made after healing and were used to search for signs of degenerative arthritis and to evaluate the lower extremity mechanical axes with respect to those on the contralateral, uninjured side.
Source of Funding
There was no external funding source for this study.
Between January 2003 and August 2006, ten patients (including seven male patients and three female patients) with diaphyseal malunions (including four femoral malunions and six tibial malunions) who met the inclusion criteria underwent a clamshell osteotomy. These procedures were completed at three separate institutions. The average age at the time of the procedure was forty-one years (range, fourteen to seventy-one years). Five patients had involvement of the right lower extremity, and five had involvement of the left. One patient had chronic obstructive pulmonary disease and hypertension, and the others had an unremarkable medical history. The mean duration of follow-up after surgery was thirty-one months (range, six to fifty-two months). All deformities were posttraumatic. The average duration of the deformity before correction was approximately eighteen years (range, one to fifty years). Treatment of previous fractures had included traction, casting, and external fixation. Three fractures were initially open. Two of the three open fractures were secondary to blunt trauma and were classified as Gustilo and Anderson type IIIA, whereas the third was a low-velocity ballistic fracture. The remaining fractures were closed and were secondary to blunt trauma. Although not an exclusion criterion, none of the patients in the present series had retention of internal fixation devices. None of the patients had had previous internal fixation, and two had been managed with external fixation devices. Three of the ten patients smoked tobacco products.
Correction of Deformity and Union
All osteotomy sites were healed clinically by the six-month follow-up visit. In addition, all but one of the osteotomy sites were thought to be healed radiographically (with bridging bone across three of four cortices) by the six-month follow-up visit. The remaining patient had bridging of two of four cortices, but no clinical symptoms, and had returned to full duty at work. In all cases, evaluation of the postoperative radiographs revealed correction of coronal and sagittal plane angulations to within 4°, complete correction of translation, and complete correction of joint-line orientation angles when compared with population averages4. Clinical evaluation revealed correction of limb-length inequalities to within 2 cm in all cases, as judged with both a tape measure and blocks for pelvic leveling. The correction of length ranged from 0 to 5 cm. The rotational profile also appeared to be fully corrected as judged on the basis of the supine resting posture of the extremity, thigh-foot axis determination in the patients with a tibial deformity, and symmetric range of rotation of the hips in both the supine and the prone position in the patients with a femoral deformity.
Hardware failure in the form of broken interlocking screws was noted in one case. One tibial nail was dynamized at four months to treat a delayed union, and the osteotomy site went on to heal uneventfully. Despite the loss of interlocking screw stability in both of these cases, loss of angular correction did not occur and one leg shortened approximately 3 mm during the healing process. One patient had postoperative delirium tremens, which resolved with symptomatic care. Two patients had a superficial wound dehiscence along the direct approach for the longitudinal component of the osteotomy. One of these patients was managed with local wound care, and the wound healed uneventfully. The second patient required operative débridement and reapproximation of the skin edges. The incision and the osteotomy sites subsequently healed without further complication. One of the two patients with wound problems smoked. No neurovascular complications were noted.
We designed this osteotomy in an attempt to simplify the correction of combined angular and translational malunions of the diaphyses of the femur and tibia. The concept of the clamshell osteotomy evolved from an extension of the experience of one of the authors (G.V.R.) with intramedullary techniques for comminuted diaphyseal fractures and knowledge of the Sofield technique5. Conceptually, the osteotomy is easy to understand. Simply stated, the clamshell osteotomy can be thought of as a comminuted diaphyseal fracture that one would treat with an intramedullary rod, with the rod acting as an anatomic axis reduction template. The concerns about length, alignment, and rotation are similar and are treated in a fashion similar to acute fracture fixation.
The present study and the described technique do have potential limitations. First, as with other observational studies, biases could be present secondary to the study design and the methods of data collection. The inclusion and exclusion criteria were set at the beginning of the study on the basis of what were thought to be appropriate indications for this technique. These criteria were very specific, and great care should be taken when evaluating patients with diaphyseal malunions who do not meet these criteria. Independent observers were not used consistently in this study to judge deformity correction and postoperative outcome scores. Although patient satisfaction was voiced when specifically sought (with one institution accounting for five patients) and radiographs revealed correction of the deformity and healing of the osteotomy sites, functional outcomes were not assessed.
Second, the osteotomy site has no inherent stability; therefore, the rod initially serves as a load-bearing device. Minimizing soft-tissue and periosteal damage is important to promote early callus formation, thereby providing implant protection through load-sharing. Despite the lack of inherent stability, partial weight-bearing was allowed immediately. To date, there has been no evidence of hardware failure leading to loss of reduction of anything other than 3 mm of length. Delayed union requiring dynamization was noted in one case. Presently, we add demineralized bone matrix to the transverse osteotomy sites if there appears to be inadequate bone contact achieved with the bone fragments produced by reaming. This situation has only occurred when lengthening was required to equalize leg lengths.
Third, certain preoperative deformity characteristics must be closely evaluated. Extensive shortening of the bone cannot be corrected safely through the malunion site in the acute setting. Correction of no more than 5 cm of shortening in the femur and 3 cm in the tibia has been attempted. Correction of shortening clearly differs from acute fracture treatment in which the soft-tissue envelope has only been shortened transiently. Appropriate caution must be taken with respect to stretching the neurovascular structures during this procedure. An understanding of the potential need for soft-tissue coverage after deformity correction should be present preoperatively. This warning is consistent with all types of acute deformity correction requiring an incision and is not specific to this technique.
Fourth, a poor soft-tissue envelope preventing a surgical approach to the malunion site is a clear contraindication to the technique. Percutaneous transverse osteotomies are possible at the extremes of the malunited segment, but the longitudinal osteotomy requires an extensile approach. This is most likely to affect the treatment of tibial deformities as the muscular envelope is more limited. Previous trauma leading to marginal soft tissues likely affected the result in one patient who initially had sustained a type-IIIA open fracture and had a postoperative wound dehiscence. Every attempt should be made to create the longitudinal incision over the anterior or superficial posterior compartment of the leg in case dehiscence does occur. An atraumatic soft-tissue approach must be a priority, and the degree of soft-tissue dissection should be minimized. On the basis of the present case series, we do not know the applicability of this technique in patients with retained or previous internal fixation.
With use of our inclusion and exclusion criteria, this osteotomy compares favorably to other techniques2,6,8,9,18. Some advantages of the osteotomy are clear: there is limited preoperative planning, and there is some leniency in the precision of the osteotomy itself. In addition, this technique can be used to manage certain patients who have an acute diaphyseal fracture proximal or distal to a diaphyseal malunion. This was successfully accomplished twice in the present series by using the acute fracture as one of the transverse limbs of the osteotomy. Our high union rate without loss of reduction, despite the creation of multiple diaphyseal healing sites, is comparable with or better than the findings associated with other published techniques2,6,8,9,18. On the other hand, problems with interlocking screw failure occurred in one patient, and problems with incision healing occurred in two patients. The need for an atraumatic soft-tissue technique is consistent with malunion correction with plate fixation8,10 but is a relative weakness when compared with correction of malunions with a multiplanar external fixator1.
One clear advantage of the fixation device is early weight-bearing, which can be permitted because of the inherent strength of the intramedullary rod. Early weight-bearing is also commonly allowed with circular frame treatment1. The real advantages of the intramedullary rod as compared with the circular frame are related to cost and the absence of pin-track concerns in association with the intramedullary rod1. Nonlocking plate fixation is another option, but early weight-bearing is usually discouraged with that technique8,9. In addition, osteotomy precision is generally more important and preoperative planning is generally more demanding when using plate fixation in order to protect the implant by creating congruent surfaces for compression.
In this small series of patients, the clamshell osteotomy proved to be a useful technique that can simplify complex lower extremity deformity correction. With appropriate patient selection, a good understanding of intramedullary rod fixation techniques, and careful soft-tissue handling, complex deformity correction can be achieved with a minimum of preoperative planning in properly selected patients.
A table showing details on all ten patients is available with the electronic versions of this article, on our web site at jbjs.org (go to the article citation and click on “Supplementary Material”) and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD).
Disclosure: The authors did not receive any outside funding or grants in support of their research for or preparation of this work. Neither they nor a member of their immediate families received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. A commercial entity (Stryker Orthopaedics) paid or directed in any one year, or agreed to pay or direct, benefits in excess of $10,000 to a research fund, foundation, division, center, clinical practice, or other charitable or nonprofit organization with which one or more of the authors, or a member of his or her immediate family, is affiliated or associated.
A video supplement to this article will be available from the Video Journal of Orthopaedics. A video clip will be available at the JBJS web site, www.jbjs.org. The Video Journal of Orthopaedics can be contacted at (805) 962-3410, web site: www.vjortho.com.
Investigation performed at University of Mississippi Medical Center, Jackson, Mississippi; University of Cincinnati Medical Center, Cincinnati, Ohio; and Harborview Medical Center, Seattle, Washington
1. Feldman DS, Shin SS, Madan S, Koval KJ. Correction of tibial malunion and nonunion with six-axis analysis deformity correction using the Taylor Spatial Frame. J Orthop Trauma. 2003;17:549-54.
2. Farquharson-Roberts MA. Corrective osteotomy for combined shortening and rotational malunion of the femur. J Bone Joint Surg Br. 1995;77:979-80.
3. Heijens E, Gladbach B, Pfeil J. Definition, quantification, and correction of translation deformities using long leg, frontal plane radiography. J Pediatr Orthop B. 1999;8:285-91.
4. Paley D, Herzenberg JE, Tetsworth K, McKie J, Bhave A. Deformity planning for frontal and sagittal plane corrective osteotomies. Orthop Clin North Am. 1994;25:425-65.
5. Sofield H, Millar EA. Fragmentation, realignment, and intramedullary rod fixation of deformities of the long bones in children. J Bone Joint Surg Am. 1959;41:1371-91.
6. Marsh DR, Shah S, Elliott J, Kurdy N. The Ilizarov method in nonunion, malunion and infection of fractures. J Bone Joint Surg Br. 1997;79:273-9.
7. Kempf I, Grosse A, Abalo C. Locked intramedullary nailing. Its application to femoral and tibial axial, rotational, lengthening, and shortening osteotomies. Clin Orthop Relat Res. 1986;212:165-73.
8. Sangeorzan BJ, Sangeorzan BP, Hansen ST Jr, Judd RP. Mathematically directed single-cut osteotomy for correction of tibial malunion. J Orthop Trauma. 1989;3:267-75.
9. Johnson EE. Multiplane correctional osteotomy of the tibia for diaphyseal malunion. Clin Orthop Relat Res. 1987;215:223-32.
10. Sanders R, Anglen JO, Mark JB. Oblique osteotomy for the correction of tibial malunion. J Bone Joint Surg Am. 1995;77:240-6.
11. Yamamoto T, Hashimoto Y, Mizuno K. Oblique wedge osteotomy for femoral diaphyseal deformity in fibrous dysplasia: a case report. Clin Orthop Relat Res. 2001;384:245-9.
12. Milch H. Osteotomy of the long bones. Springfield: Charles C Thomas; 1947.
13. Mast JW, Teitge RA, Gowda M. Preoperative planning for the treatment of nonunions and the correction of malunions of the long bones. Orthop Clin North Am. 1990;21:693-714.
14. Hierholzer G. Indications for corrective osteotomy after malunited fractures. In: Hierholzer G, editor. Corrective osteotomies of the lower extremity after trauma. Berlin: Springer-Verlag; 1985. p 9-28.
15. Freedman EL, Johnson EE. Radiographic analysis of tibial fracture malalignment following intramedullary nailing. Clin Orthop Relat Res. 1995;315:25-33.
16. McConnell T, Tornetta P 3rd, Tilzey J, Casey D. Tibial portal placement: the radiographic correlate of the anatomic safe zone. J Orthop Trauma. 2001;15:207-9.
17. Sagi HC, Papp S, Dipasquale T. The effect of suture pattern and tension on cutaneous blood flow as assessed by laser Doppler flowmetry in a pig model. J Orthop Trauma. 2008;22:171-5.
18. Mayo KA, Benirschke SK. Treatment of tibial malunions and nonunions with reamed intramedullary nails. Orthop Clin North Am. 1990;21:715-24.