Intramedullary Nail-Guided Correction of Lower Extremity Long Bone Deformity: Technique and Case Series

1. INTRODUCTION

Long bone deformity is commonly seen in patients with metabolic bone disease, congenital deformity, growth disturbance, or impaired osteogenesis following traumatic injury. In the lower limb, deformity may lead to pain, fractures, and impaired function [1-3]. The natural history of untreated deformities often includes early arthrosis due to an abnormal mechanical axis [4-9]. Surgical intervention aims to correct the mechanical axis of the lower extremity, leading to a lower risk of subsequent fracture and arthrosis, as well as aiding patient mobility and independence. Traditionally, treatment of long bone deformity with external fixation has achieved positive clinical outcomes via gradual correction. However, this demands a significant financial and time commitment from both patient and provider. Additionally, external fixators carry a risk of pin-site infections, and prolonged soft-tissue transfixation can cause scarring and pain and may even impair the range of motion, affecting postoperative rehabilitation [10-12].

More recently, computer-assisted hexapod external fixators have been used in Deformity correction (DC) [13-15]. This type of frame (two or more rings, connected by six telescopic struts) allows for precise three-dimensional DC, as well as for gradual correction over time. One of this technology’s main advantages is that it allows the surgeon to input defined parameters for the current deformed bone state, as well as the desired correction, leaving the software to determine precise strut adjustments required to accomplish the correction. This use of computerized software to assist in complex DC has been further expanded to assist with acute DC. Authors have described the use of Computer Hexapod-Assisted Orthopedic Surgery (CHAOS) as an alternative method to achieve precise DC intraoperatively [12,16,17]. CHAOS applies a hexapod external fixator only intraoperatively, utilizing strut settings and computer software to assist with DC. After the desired DC is achieved, definitive internal fixation via plate or intramedullary nail (IMN) is performed, with removal of the hexapod frame intraoperatively. Proponents of CHAOS indicate that it enables precise, minimally invasive multiplanar DC intraoperatively. Multiple studies report highly accurate correction with CHAOS for a variety of lower extremity deformities, accompanied by a low complication rate [12,16-18].

However, utilization of the hexapod external fixator is both costly and time-intensive. The quoted cost of the frame, struts, and necessary pins can vary, but estimates are as high as $10,675 [19]. Procedure duration is also substantially increased with CHAOS, with an average of 43 minutes required for frame application intraoperatively prior to beginning internal fixation [12].

The aim of this retrospective cohort study was to assess the efficacy of intramedullary nail-guiding (IMNG) as a non-inferior cost-effective alternative approach to CHAOS for lower extremity long bone DC. It is hypothesized that the IMNG approach can achieve DC at a rate of complications equal to or lower than that of the CHAOS procedure. Surgical description of the IMNG technique is supplemented by a review of clinical and radiographic outcomes of the technique in a patient cohort treated at a single institution.

2. METHODS

2.3. Operative Technique - Femoral Deformity

Deformities proximal to the isthmus were often managed with an antegrade femoral nail, while deformities of the distal femur were corrected with a retrograde IMN. In the case of antegrade femoral nailing, the guidewire is fluoroscopically positioned onto the appropriate starting point of the femur, based on the selected nail geometry. The type of nail, i.e., trochanteric versus piriformis, is less important than understanding the exact geometry of the nail and its relationship to the anatomy of the patient’s proximal femur. Templating software is utilized to ensure the paths of the opening reamer and IMN through the proximal femur are exactly colinear to the anatomic axis of that bone segment. After opening reaming, the intramedullary reduction tool is inserted into the femur and utilized as a surrogate for eventual nail placement. Based on preoperative planning, standard wedge osteotomy is carried out at the apex of the deformity.

For distal femoral deformities, a retrograde nail combined with supracondylar dome osteotomy is typically used. In this case, a longitudinal incision is made just proximal to the patella, splitting the quadriceps tendon in line with its fibers, to access the anterior supracondylar femur, and the dome osteotomy is performed. A single guide pin is placed at the apex of the deformity, marking the point around which the dome osteotomy should rotate. Using a parallel pin guide, multiple drill holes are made in a dome-shaped pattern and subsequently connected using an osteotome. To access the starting point for the retrograde nail, a mini-medial parapatellar approach is made just proximal to the tibial tuberosity. After dissecting through the skin and subcutaneous tissue, the patellar tendon is identified and retracted laterally. The entry portal for the retrograde IMN at the intercondylar notch is identified fluoroscopically. Fluoroscopy is used to confirm the appropriate starting point and trajectory that is exactly colinear with the anatomic axis of the distal femur. The guide pin is placed, and the cannulated opening reamer is utilized to create the initial nail pathway through the distal femur. Based on patient anatomy, an appropriately sized IMN is advanced to the point of the osteotomy. Appropriate nail positioning is confirmed via fluoroscopy throughout advancement. At the osteotomy site, the bone is manipulated to align the nail with the anatomical axis, and the nail is advanced further, correcting the deformity. After advancement of the nail, satisfactory reduction and restoration of the anatomic and mechanical axes are again confirmed fluoroscopically. The nail is then locked either proximally or distally, the osteotomy site is compressed, and the nail is secured proximally with 2 statically locked bolts. If a closing wedge osteotomy was performed, the wedge of bone is excised, any cancellous bone removed is packed back into the osteotomy site, and a standard closure is performed.

Details regarding surgical technique are supplemented by an example of closing wedge osteotomy of a distal femoral deformity (Fig. 3).

3. RESULTS

Twenty-four IMNG procedures were performed in 18 patients, 10 of whom were female. At the time of surgery, the mean patient age was 30 (15 to 62) years, and the mean body mass index was 31 (21-67) kg/m². Ten (56%) patients had metabolic bone disease, 4 (22%) had a history of congenital deformity or childhood growth disturbance, and 3 (17%) patients required DC for malunion following trauma (Table 1). The mean time of follow-up after surgery was 36 months (range 9 to 111 months). Eight patients underwent correction of isolated tibial deformities, of whom two had contralateral tibial deformities corrected in separate operations. Ten patients underwent isolated femoral DC, and two patients also underwent multiple DC operations for both femoral deformity and tibial deformity.

Five patients had pre-existing unicompartmental osteoarthritis with a medially targeted mechanical axis. Patient 18 was indicated for staged knee arthroplasty following femoral deformity correction and, as such, was templated to have a residual valgus deformity to accommodate the anticipated total knee components.

Pre- and postoperative radiographs were available for 16/18 (89%) patients to assess changes in standard lower limb angles post-correction. The mean preoperative MAD was 42.5 mm, ranging from 3 mm to 134 mm away from the targeted postoperative MAD. The patient with preoperative MAD of 3 mm had a severe varus deformity of the femur. Postoperative average MAD was 10.2 ± 4.8 mm from the targeted MAD, with overall MAD ranging from 0.55 to 14.9 mm from the target. In patients with femoral deformities, preoperative mechanical and anatomic LDFAs averaged 87.6° (range: 74° to 111°) and 77.9° (66.6° to 100°), respectively. Postoperative mechanical and anatomic LDFAs were 89.3° (85.7° to 99°) and 82.5° (76.7° to 92.3°), respectively. In patients with tibial deformities, the average preoperative MPTA was 84.8° (77.4° to 94.1°), and postoperative MPTA was 86° (73.1° to 92.4°). LDTA averaged 85.9° (50.6° to 96.8°) preoperatively as compared with 87.1° (73.6° to 93.6°) postoperatively (Table 2).

Definitive fixation was achieved in each case with a statically locked IMN. Twelve procedures were performed for uniplanar DC, and ten procedures corrected biplanar deformities. Seven patients had two-level osteotomies. All corrections of tibial deformities were performed with closing wedge osteotomies. Femoral deformities were corrected with closing wedge osteotomies in five procedures and dome osteotomies in five procedures. Femoral corrections were performed with retrograde IMN in 8 cases and anterograde IMN in 4 cases. Postoperative rehabilitation protocol encouraged immediate weight-bearing as tolerated in the operative extremity.

Complications occurred in 5 patients with tibial DC, including 3 cases of hardware removal due to symptomatic prominent interlocking screws and 1 case of delayed union. Complications arose in 4 patients with femoral DC, including 3 cases of nonunion, 1 case of hardware removal due to prominent screws, and 1 case of nonunion and infection (Table 1). There were no cases of peroneal nerve palsy or compartment syndrome. All patients achieved eventual clinical and radiographic union. Pre- and postoperative radiographs of 4 cases are included to highlight the variation in presentation and operative site (Figs. 4-7).

4. DISCUSSION

Results of this patient series demonstrate that the IMNG approach to complex tibial and femoral deformities is reliable and effective at correcting large deviations in limb alignment. The patients in this cohort had a high degree of variation in deformity types and preoperative MAD, including an MAD that measured 134 mm in one patient. Postoperative MAD was corrected to an average of 10.2 mm from the target, demonstrating that IMNG is well-suited for realigning the lower extremity biomechanical axis within the normal range of MAD, previously defined by Paley et al. as 9.7 ± 6.8 mm medial to the center of the knee joint line [20].

Previous studies conducted on patients who underwent DC with the CHAOS technique report restoration of anatomic alignment, though there has been variation in the data points reported. In the initial CHAOS paper by Rogers et al., MAD was represented as a ratio of distance from the center of the knee joint to the width of the tibial plateau, rather than measured in millimeters, to account for variation in the size of radiographs available [12]. Though an MAD ratio correction of 0.42 and a postoperative MAD ratio of 0.11 were reported, it is difficult to quantify the exact postoperative MAD. In a 2016 study by Hughes et al. on tibial DC in 12 patients using CHAOS, postoperative tibial mechanical axis was reported as 0.7° varus, which is in line with historical normal anatomical measurements [17,21,22]. In the most recent CHAOS study to report MAD, French et al. reported mean postoperative MADs of 4 mm and 7.4 mm for valgus and varus deformities, respectively, though the maximum MAD postoperatively ranged up to 27 mm[18]. In comparison to previous studies on CHAOS, IMNG demonstrates a similar level of precision in overall DC as measured by postoperative MAD.

Compared with CHAOS, IMNG is accompanied by a comparable or lower rate of complications, especially relating to soft tissues. The additional pin-sites necessary for hexapod frame attachment to various segments of the deformed bone expose the patient to a greater risk of soft tissue injury or infection, especially in the setting of multilevel osteotomies. Compared with IMNG, which resulted in a relatively low infection rate (5.6%) and did not result in any nerve injuries or vascular complications, CHAOS studies report varying soft tissue complication rates ranging up to 33%, including common peroneal nerve palsies, tibial pseudoaneurysms, and superficial wound infections [12,16-18]. The rate of nonunion or delayed healing in this study was 22%, which is slightly higher compared to the rate of nonunion and malunion with CHAOS (0 - 14%). The rate of hardware removal in patients treated with IMNG was 22%; in the two largest cohorts of patients treated with CHAOS, subsequent hardware removal was required in 41 - 58% of cases, though the studies did not further delineate which of these patients required removal for soft tissue irritation versus secondary surgery [16,18]. Nevertheless, the rate of hardware removal in patients in this IMNG cohort is approximately half of that reported with CHAOS and is an important consideration when reducing the overall morbidity associated with DC.

Perhaps the most significant advantages of the IMNG approach are the simplicity of the surgical technique and decreased financial burden. Surgical procedures are a major expense for the healthcare system, and operating room costs make up the second-largest portion of surgical care spending [23]. Review of the manufacturer’s suggested retail price of one example hexapod frame system (Taylor Spatial Frame, Smith and Nephew, Memphis, TN) compared with intramedullary nailing with interlocking screws revealed a four-fold difference in cost; the hexapod frame retail price is approximately $39,000, whereas the cost for IMNG is roughly $10,000. This gap in price widens further after accounting for the fact that 20-33% of patients undergoing CHAOS required multilevel osteotomies, necessitating additional hexapod frames [16,18]. However, it must be noted that these price estimates assume the hexapod frame to be a single-use item, which has been reinforced in existing literature [16,18]. At the study institution, reuse of any part of the hexapod frame system is similarly prohibited. This discrepancy in cost may be potentially reduced at certain institutions that permit the reuse of struts or ring components for temporary intraoperative use. Institutional single-use versus reuse of hexapod frame system components is not well-described, and data from other institutions are needed to better characterize variation in equipment expenses. Importantly, the savings in operating room time with IMNG in comparison to CHAOS are also substantial. Rogers et al. reported that application of the hexapod frame intraoperatively required between 35 and 60 minutes prior to initiation of internal fixation [12]. A recent study by Moody et al. on the cost of orthopedic trauma surgery operating time reported an average cost of $16 per minute, which can translate to roughly $1,000 in additional cost solely from pre-fixation frame application time [24]. While this study does not report the exact time savings that the IMNG approach affords, this technique definitely requires fewer time and steps than CHAOS.

This study is limited by its retrospective nature and its small sample size. Furthermore, because not all patients in this cohort had hip-to-ankle standing radiographs available for review, some anatomical measurements, such as LDFA, MPTA, and LDTA, were recorded using long bone radiographs. Given that most patients did not have lateral hip-to-ankle standing radiographs postoperatively, it was not possible to quantify the degree of correction of procurvatum and recurvatum deformities. Future studies should consider standardizing postoperative lateral hip-to-ankle standing films, in addition to anteroposterior, for review by the orthopedic surgeon to better assess these deformities.

CONCLUSION

This study demonstrates accurate correction of complex multiplanar tibial and femoral deformities using the IMNG approach, while maintaining a low rate of complications in comparison to other DC techniques. IMNG should be considered a cost-efficient correction option with readily available implants for restoring alignment and function in patients with complex deformities of the femur and tibia.

REFERENCES