If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Charcot medial column and midfoot deformities are associated with rocker bottom foot, recurrent plantar ulceration, and consequent infection. The primary goal of surgical intervention is to realign and stabilize the plantar arch in a shoe-able, plantigrade alignment. Different fixation devices, including screws, plates, and external fixators, can be used to stabilize the Charcot foot; however, each of these methods has substantial disadvantages. To assess the effectiveness of rigid, minimally invasive fixation of the medial column and midfoot, 8 cases of solid intramedullary bolt fixation for symptomatic Charcot neuroarthropathy were reviewed. The patients included 6 males (75%) and 2 females (25%), with a mean age of 63 (range 46 to 80) years. The Charcot foot deformity was caused by diabetic neuropathy in 7 cases (87.5%) and alcoholic neuropathy in 1 (12.5%). The mean duration of postoperative follow-up period was 27 (range 12 to 44) months. The mean radiographic correction of the lateral talar–first metatarsal angle was 15° (range 3° to 19°), and the mean radiographic correction of the dorsal midfoot dislocation was 9 (range −4 to 23) mm. The mean loss of correction of the lateral talar–first metatarsal angle and midfoot dislocation after surgery was 7° (range 0° to 26°) and 1 (range 0 to 7) mm, respectively. No bolt breakage was observed, and no cases of recurrent or residual ulceration occurred during the observation period. Bolt removal was performed in 3 cases (37.5%), 2 (25%) because of axial migration of the bolt into the ankle joint and 1 (12.5%) because of infection. The results of the present review suggest that a solid intramedullary bolt provides reasonable fixation for realignment of the medial column in cases of Charcot neuroarthropathy.
Charcot neuroarthropathy commonly causes chronic destructive arthropathy involving the joints of the foot and ankle. It occurs as a consequence of various peripheral neuropathies, most commonly as a result of diabetes mellitus (
). Typically, destruction of the Lisfranc joint complex causes flattening of the longitudinal foot arch (rocker bottom deformity), leading to plantar ulceration, which, especially in the course of diabetes, often results in infection and significant morbidity.
When conservative management fails to heal or prevent recurrent ulceration, surgical treatment options must be considered (
). Minimizing soft tissue damage during surgery is essential, because incomplete or slow healing is generally known to be associated with neuropathy, vasculopathy, dermopathy, and difficulty avoiding weightbearing activities, all of which are known to affect this patient population. Rigid osteosynthesis is also important owing to the high likelihood of uncontrolled postoperative weightbearing, which can lead to failure of the implants and reconstruction (
Several different types of bone fixation devices, including screws, plates, and external fixators, can be used to stabilize the reconstructed Charcot foot; however, each of these devices has substantial disadvantages that can affect the clinical course. Cannulated screws break easily, and although a combination of plates and screws is stronger than screws alone (
); however, the risk of pin site infection remains, and the fixators can be inconvenient to patients.
Recently, several investigators have proposed restoring the alignment of the medial column (medial metatarsocuneiform, medial naviculocuneiform, and talonavicular joints) by intramedullary placement of a cannulated screw (
). The clinical results have been favorable; however, screw breakage and migration is a substantial risk associated with this form of fixation owing to the decreased relative strength of a cannulated screw compared with a solid-core screw. With screw breakage, loss of alignment can occur, and additional surgical intervention can become necessary. In an effort to assess the clinical and radiographic outcomes of reconstruction of the Charcot foot using a solid fusion bolt for intramedullary beam stabilization of the medial column and midfoot, we undertook a review of 8 feet in 8 patients with neuroarthropathy and symptomatic rocker bottom foot.
Patients and Methods
We conducted a review of consecutive patients who were seen in our clinic and treated for Charcot neuroarthropathy-associated midfoot collapse using midfoot fusion and a solid intramedullary fusion bolt. The regional ethics committed approved the investigation, and written informed consent for inclusion in the review was obtained. The medical records were reviewed by 1 of us (M.M.), who searched for patients using the diagnosis code for Charcot foot and the diagnosis code for arthropathy associated with other endocrine and metabolic disorders (International Classification of Diseases, 9th Revision, codes 713.5 and 713.0, respectively, World Health Organization, Geneva, Switzerland). To be included in the case series, all that was required was the diagnosis of Charcot foot and treatment using intramedullary beaming and the solid fusion bolt, along with preoperative and postoperative radiographs of the involved foot.
Alignment of the medial column and midfoot, specifically the medial metatarsocuneiform, medial naviculocuneiform, and talonavicular joints, was restored and fixed using intramedullary placement of a 6.5-mm diameter solid bolt (Midfoot Fusion Bolt, Synthes, Zuchwil, Switzerland) (
). Joint realignment was performed with a full, open dissection. Joint resection of the medial column in preparation for fusion was performed (cases 1 through 4 and 8); alternatively, the joint surfaces where left intact (cases 5 through 7). The intramedullary bolt was inserted in either a retrograde fashion, directed from the distal end of the first metatarsal into the talus, or an antegrade fashion, directed from the posterolateral aspect of the talus into the first metatarsal. Additional midfoot joints were fused, if needed, using additional large diameter screws or other forms of internal fixation. If the Achilles tendon was found to be tight, effecting ankle equinus, the aponeurosis of the gastrocnemius or the Achilles tendon was lengthened before bone realignment. Postoperative care consisted of the use of a non–weightbearing orthosis for the first 12 weeks, followed by gradual progression to full weightbearing according to the clinical and radiographic findings.
The medical records were reviewed, and the clinical history information was abstracted. The clinical examination at the last follow-up visit included the American Orthopaedic Foot and Ankle Society midfoot score (
). The Inlow screening score reflects the residual risk of foot ulceration and ranges from 0 (lowest risk) to 23 (highest risk). We were also interested in whether the patients could ambulate with supportive shoe gear with or without a cane after surgery and healing.
All patients underwent pre- and postoperative radiographic imaging. The patterns of osseous deformity as described by Sammarco et al (
). Radiographic measurements were taken with the patient weightbearing in the preoperative setting, non–weightbearing in the immediate postoperative setting, and weightbearing at the last follow-up evaluation. The talar–first metatarsal angle was measured in the anteroposterior view, with positive value representing valgus deformity (
). A paired Student's t-test was performed to compare the radiographic measurements from the pre- and postoperative periods, with statistical significance defined at the 5% (p ≤ .05) level.
From October 2007 to July 2010, we performed 10 unilateral midfoot arthrodeses in 10 patients using solid intramedullary bolt fixation of the medial column and midfoot. Two patients were lost to follow-up because they died, which was determined to be unrelated to their foot operation. We reviewed the clinical and radiographic data pertaining to the 8 remaining cases, and the demographic data are listed in Table 1. The study cohort consisted of 6 males (75%) and 2 females (25%), and their mean age was 63 (range 46 to 80) years. The mean postoperative follow-up duration was 27 (range 12 to 44) months. Of the 8 patients, 7 (87.5%) had diabetes mellitus, and 1 (12.5%) had alcoholic neuropathy. Of the 8 feet, 3 (37.5%) were stabilized with the bolt inserted in an antegrade fashion, and 5 (62.5%) were stabilized in a retrograde fashion. At surgery, 4 patients (50%) had a plantar ulcer. According to the Wagner ulcer grading system (
), 3 (37.5%) of the ulcers were grade 3 and 1 (12.5%) was grade 2. Of the 8 feet, 3 (37.5%) had a Sammarco grade 3 Charcot foot deformation, 3 (37.5%) had grade 1 deformation, and 1 (12.5%) had Sammarco grade 5 deformation (
), 3 (37.5%) were type I, 3 (37.5%) were type III, and 1 (12.5%) was type IV. Five of the feet (62.5%) underwent concomitant fusion of other joints in the same foot. As can be seen in the radiographs (Figs. 1 and 2), stability after bolt insertion need not be associated with complete osseous fusion of the splinted joints.
Table 1Patient characteristics (n = 8 feet in 8 patients)
Type I, LisFranc's joint; type II, naviculocuneiform; type III, perinavicular; type IV, Chopart's joint; subclasses: A, above metatarsal–calcaneus plane; B, coplanar; C, below; a β stage is assigned if 1 of the following criteria is met: dislocation is present, the lateral talar–first metatarsal angle is ≥30°, the lateral calcaneal–fifth metatarsal angle is ≥0, or the anteroposterior talar–first metatarsal angle is ≥35°; an α stage can be assigned when all 4 features are absent.
First and second metatarsal, medial–intermediate cuneiform (plates)
First and second metatarsal, second, third, and fourth TMT, medial–intermediate cuneiform (screws)
First and second metatarsal, second and third TMT (screws)
First and second metatarsal, third TMT (screws)
Abbreviations: NA, not available; TMT, tarsometatarsal joint.
Data in parentheses are ranges.
∗ Type I, LisFranc's joint; type II, naviculocuneiform; type III, perinavicular; type IV, Chopart's joint; subclasses: A, above metatarsal–calcaneus plane; B, coplanar; C, below; a β stage is assigned if 1 of the following criteria is met: dislocation is present, the lateral talar–first metatarsal angle is ≥30°, the lateral calcaneal–fifth metatarsal angle is ≥0, or the anteroposterior talar–first metatarsal angle is ≥35°; an α stage can be assigned when all 4 features are absent.
The clinical and radiographic data are presented in Tables 2 and 3 and Fig. 3. No cases resulted in limb amputation, and all the patients were able to walk at least 1 block, with a cane in 2 (25%) and without a cane in 6 (75%) at the last follow-up evaluation. The mean American Orthopaedic Foot and Ankle Society midfoot score (
) at the final follow-up evaluation was 67 (range 58 to 83). In addition, at the final follow-up visit, no cases of recurrent or residual ulceration had developed. The mean Inlow 60-Second Diabetic Foot Screen (
) score was 8 (range 5 to 11). Preoperative conventional radiographs were unavailable for 1 patient (12.5%; patient 2). Bolt migration occurred in a total of 3 cases (cases 5 through 7). Bolt removal became necessary in 3 patients (37.5%), 1 (12.5%) because of a deep infection followed by massive osteolysis and collapse of bone structure (patient 8) and 2 (25%) because of axial migration of the bolt into the ankle (patients 6 and 7). None of the bolts deformed or broke.
Table 2Postoperative clinical and radiologic data at last follow-up visit
The amount of surgical correction was determined by comparing the preoperative weightbearing and immediate postoperative non–weightbearing radiographs taken immediately after surgery (Table 3). The mean correction values were as follows: anteroposterior talar–first metatarsal angle, 12° (range 2° to 30°); lateral talar–first metatarsal angle, 15° (range 3° to 21°); lateral calcaneal–fifth metatarsal angle, 11° (range 4° to 20°); and midfoot dorsal displacement, 9 (range −4 to 23) mm. These differences were all statistically significant (p < .05). Maintenance of the correction was determined by comparing the immediate postoperative non–weightbearing and long-term postoperative weightbearing foot radiographs (Table 3). The mean loss of correction was as follows: anteroposterior talar–first metatarsal angle, 2° (range −3° to 7°), lateral talar–first metatarsal angle, 7° (range 0° to 26°); lateral calcaneal–fifth metatarsal angle, 9° (range 1° to 16°); and midfoot dorsal displacement 1 (range 0 to 7) mm. A statistically significant (p < .05) postoperative decrease was seen in the lateral calcaneal–fifth metatarsal angle (p ≤ .05) but not in the other 3 measurements. A representative case of successful midfoot fusion with a solid bolt (patient 3) is shown in Fig. 1, and a case in which the bolt migrated and required subsequent removal (patient 7) is shown in Fig. 2.
Charcot midfoot deformity is associated with recurrent plantar ulceration and consequent morbidity. The aim of surgery is, therefore, to achieve a plantigrade foot with normal plantar pressure distribution (
) reported that nonoperative treatment is associated with an approximately 2.7% annual rate of amputation, a 23% risk of requiring bracing for more than 18 months, and a 49% risk of recurrent ulceration.
Surgical intervention to restore the plantar arch is typically a last resort option for patients in whom conservative treatment is not feasible or has failed. Therefore, patients undergoing reconstructive surgery usually present with a severe deformity. In our case series, the mean preoperative radiologic lateral talar–first metatarsal angle was −18°, and the mean midfoot dorsal dislocation was 11 mm. Such severe deformities require a substantial amount of realignment to effect satisfactory correction. In our case series, the mean correction of the lateral talar–first metatarsal angle and midfoot dorsal displacement angle was 15° and 9 mm, respectively. As such, a firm fixation method was necessary to try and maintain this amount of correction. Because patients with Charcot foot cannot control the weightbearing load owing to a lack of sensation and other morbidities (e.g., obesity, concomitant arthritis, visual deficits, cardiovascular disease), repetitive peak loads can lead to implant failure.
To secure the realignment of the Charcot foot, skeletal fixation with screws, plates, and external fixation have been used (
) demonstrated the biomechanical superiority of plantar plates for midfoot stabilization in a cadaver study. However, placement of a plate usually requires extensive dissection and exposure, resulting in accompanying soft tissue damage. Additionally, the implant construct can be bulky and increase tension in the overlying skin. External fixation has been recommended by several investigators (
) reported a series of 51 such patients with a minimum follow-up period of 1 year, in which the mean lateral talar–first metatarsal angle was reduced from 27.6° to 6.4°. However, external fixation is known to be associated with a substantial risk of pin site infection and patient inconvenience owing to the prominence of the frame.
) reviewed 22 cases of midfoot arthrodesis with a cannulated screw and reported significant reduction and maintenance of the talar–first metatarsal and calcaneal–fifth metatarsal angles after a mean follow-up period of 52 months. The key advantage of this method is that a long incision and wide exposure is not required for fixation. However, the implants usually used for such procedures are large-diameter cannulated screws, and it has been shown that these screws carry a substantial risk of breakage. Sammarco et al (
) used 6.5-mm cannulated screws and reported breakage of the screws in 8 (36.4%) of 22 cases.
The main form of fixation used in the present patients was a 6.5-mm diameter, solid, headless bolt. At a mean follow-up duration of 27 months, no bolt breakage had occurred, and the midfoot alignment was maintained in all but 1 patient, in whom postoperative infection caused massive osteolysis and failure (patient 8). In our case series, correction of the calcaneal–fifth metatarsal angle was not maintained at the last follow-up examination. We believe this was because we did not always beam the lateral column, and the radiographic angle was determined using the non–weightbearing radiographs immediately after surgery. We believe that the lateral column fixation is, however, less important for favorable clinical results, although a longer follow-up period is needed to determine whether this theory is likely to be valid. Because 1 of the most important aims of realignment surgery for patients with Charcot foot is to avoid recurrent ulceration, and none of the patients in our series experienced postoperative ulceration or amputation during the observation period, it seems that beam stabilization of the medial column and midfoot is clinically beneficial, and the residual risk of foot ulceration is low according to the Inlow 60-Second Diabetic Screen.
A drawback of the fusion bolt used in our patients is the predilection for axial migration, which we observed in 3 patients (37.5%). In 2 of those patients, the midfoot bolt penetrated into the ankle joint and necessitated subsequent implant removal. Interestingly, bolt migration occurred only in cases in which the intramedullary bolt was used to stabilize the medial column without necessarily aiming to fuse (no joint surface resection). We believe the reason for this was the short and shallow thread of the bolt, poor bone stock in patients with Charcot disease, and subsequent weightbearing on a foot that had not, in all cases, achieved solid, radiographic evidence of arthrodesis along the beamed column of bones. We hypothesized that changing the design of the bolt to 1 with a deeper and longer thread pattern might decrease the possibility of axial bolt migration. Also, the use of an interlocking mechanism, similar to that used in femoral and tibial intramedullary nails, might also minimize the risk of this complication.
As with all retrospective case series, we appreciate that numerous methodologic limitations threaten the validity of any conclusions we have made. For instance, coding biases could have influenced our ability to identify potentially eligible patients, and we omitted 2 patients who had undergone the procedure but died of causes determined to be unrelated to Charcot foot surgery. Similarly, the retrospective design limited our ability to undertake an explanatory analysis, and we could not determine whether the comorbidities such as renal disease, obesity, or any other factors influenced our outcomes. Furthermore, we could not compare the pre- and postoperative clinical outcomes meaningfully. Still further, our emphasis was on radiographic measurement suggestive of a more normal pedal alignment, although the precise meaning of such measurements, whether weightbearing or not, in regard to subjective patient satisfaction is not known. We were also unable to determine the association of solid radiographic evidence of fusion with satisfactory long-term outcomes compared with the influence of beam stabilization without radiographic arthrodesis. Despite these shortcomings, we believe the results from the present series of patients imply that the solid fusion bolt can be used to stabilize the realigned foot in most patients with symptomatic medial column and midfoot Charcot deformation.
In conclusion, the results of the present preliminary series of patients could be useful in the design of future prospective cohort studies and randomized controlled trials focusing on surgical repair of the Charcot foot.