Foot & Ankle International

Revision surgery in the presence of infection carries high risks. We describe our results using a new technique to treat these challenging problems. We treated infected nonunions with cavitary voids with adjuvant antibiotic loaded calcium sulfate–hydroxyapatite paste composite and autologous bone graft (ABG) layer technique coupled with stable fixation.

Thirty consecutive patients who underwent revision foot and ankle surgery for an infected nonunion were prospectively studied. Following multidisciplinary team workup, surgical debridement and biopsies were undertaken. Bone voids were measured and classified according to containment and size. ABG was mixed and layered with an adjuvant antibiotic-loaded calcium sulfate–hydroxyapatite paste followed by surgical reconstruction including arthrodesis and fixation. Empirical and pathogen-specific antibiotics were instituted until intraoperative sample-specific antibiotics were identified and used. Patients were prospectively followed up for a minimum of 1 year.

The male-female ratio was 16:14, mean age was 51.3 years, and 23.3% smoked at definitive surgery. Void volume was <1 cm³ (n=9), 1-2 cm³ (n = 13), and >2 cm³ (n=8). No patients either were lost to follow-up or had a further infective episode at a mean of 38.3 months; 86.7% united with fusion on imaging. Four patients had radiographic evidence of nonunion; 3 were asymptomatic and 1 required revision surgery (void >2 cm³). Independent ambulation was achieved at an average of 12 weeks, at 1 year mean American Orthopaedic Foot & Ankle Society score was 77.7 (SD 9.59), and the Manchester-Oxford Foot Questionnaire reached an effect size >0.5 in all domains at 1 year following surgery. The union rate was independent of smoking status and vitamin D deficiency (P = .94).

Layered autologous bone grafting with adjuvant antibiotic-loaded calcium sulfate–hydroxyapatite paste has been shown to be effective and safe in revision arthrodesis, with low comorbidities in void gaps without infection recurrence.

Bunionette deformity is described as a painful bony prominence on the lateral aspect of the fifth metatarsal head. The present study prospectively assessed the clinical, functional, and radiographic outcomes of sliding distal metatarsal minimally invasive osteotomy (S-DMMO) used in a large series of patients with a midterm follow-up period.

From December 2015 to December 2018, we evaluated 74 feet (57 patients). Radiologic (4-to-5 intermetatarsal angle, fifth metatarsophalangeal angle, the fifth metatarsal head width, lateral deviation angle, fifth metatarsal length, medial displacement, and elevation), clinician-reported scores (AOFAS score, visual analog scale [VAS]) and patient-reported outcomes measure (Foot and Ankle Ability Measure [FAAM] activities of daily living [ADL], FAAM sports, Manchester-Oxford Foot Questionnaire [MOXFQ], and patient satisfaction survey) were included in the analysis. The time to bone union was also assessed.

The average 4-to-5 intermetatarsal angle improved from 11.1 degrees preoperatively to 4.5 degrees postoperatively (P < .001), whereas the average fifth metatarsophalangeal angle improved from 15.7 degrees preoperatively to 4.8 degrees postoperatively (P < .001). The lateral deviation angle was found to be 2.2 degrees. The fifth metatarsal length decreased from 66.6 to 64.3 mm postoperatively. The average initial medial displacement was 4.67 mm preoperatively and 4.54 mm at final follow-up. The elevation of the fifth metatarsal head was 1.08 mm. The American Orthopaedic Foot & Ankle Society score improved from 54.3 ± 20.86 points preoperatively to 93.4 ± 17.3 (P < .001). VAS score decreased from 7.9 to 0.7 (P < .001). Also, FAAM ALD, FAAM Sport, and MOXFQ showed statistically significant differences (P < .001) between preoperative and postoperative periods. Patients found the procedure excellent in 89.1% of cases, very good in 5.4%, and good in 5.4%. Bone union was obtained at 8.1 weeks.

The most important finding of the present study is that S-DMMO showed improvement in function and pain associated with a high satisfaction rate. Also, we found substantial capacity to correct deformities and a low incidence of complications.

Level IV, case series.

Classification of fifth metatarsal base fractures has been a source of confusion since originally described by Jones in 1902. Zone classifications have been described but never evaluated for reliability. The most recent classification, metaphyseal vs meta-diaphyseal, may be unknown to many surgeons. The purpose of this study was to evaluate reliability of American Orthopaedic Foot & Ankle Society (AOFAS) members classifying fifth metatarsal base fractures and current management of these fractures.

A survey was emailed to AOFAS members including radiographs of 18 fifth metatarsal base fractures. Demographic information was collected in addition to evaluation of the radiographs. Interrater reliability was assessed for each measurement: presence of Jones fracture, zone classification, and metaphyseal vs metaphyseal-diaphyseal, using Fleiss kappa. After 3 weeks, a second email was sent to the members asking to retake the survey to evaluate intrarater reliability. Respondents were asked which region is a Jones fracture, which classification is used, if symptomatic zone 2 and 3 fractures are treated similarly, and what fractures are operative in healthy symptomatic acute fractures.

A total of 223 AOFAS members, with a median time in practice of 12 years (range 0-50), completed the initial survey. Eighty members (36%) repeated the survey for intrarater comparison. Interrater reliability was moderate for Jones and zone classification but substantial for the 2-zone metaphyseal/meta-diaphyseal classification. The median intrarater kappa was 0.78, 0.75, and 0.78 for Jones, zone, and metaphyseal/meta-diaphyseal respectively. Seventy percent of respondents treat zones 2 and 3 similarly, and approximately 60% consider an acute symptomatic fracture identified as Jones, zone 2 or zone 3 operative.

A 2-zone system may be the best available classification for fifth metatarsal base fractures given high interrater reliability and 70% of AOFAS members treat zones 2 and 3 in similar fashion.

Level III, diagnostic study.

We previously reported an increase in pronation of the first metatarsal (M1) head relative to the ground in hallux valgus (HV) patients compared to controls. Still, the origin and location of this hyperpronation along the medial column is unknown. Recent studies showed that presence of progressive collapsing foot deformities (PCFDs), which is a condition frequently associated with HV, can strongly influence the medial column coronal plane alignment. The objective of this study was to assess the coronal rotation of the medial column bones in HV feet, HV feet with radiologic markers of PCFD, and controls. We hypothesized that hyperpronation in HV will originate from a combination of M1 intrinsic torsion and first tarsometatarsal joint malposition.

The same cohort of 36 HV and 20 controls matched on age, gender, and body mass index was used. Previously, a validation of the measurements was carried out through a cadaveric study. Using these metrics, we assessed the coronal plane rotation of the navicular, medial cuneiform, and the M1 at its base and head with respect to the ground using weightbearing CT images. We measured the Meary angle and the calcaneal moment arm in our 36 HV subjects. We subdivided our cohort into an HV group and a potential PCFD HV group according to these measurements. Comparisons on medial column bones coronal rotation were performed between HV, PCFD HV, and control groups.

Twenty-two HV cases were included in the HV group and 14 in the PCFD HV group. Both groups presented an increase in pronation of the first metatarsal head relative to the ground when compared to the control group (P < .001). Comparing HV and controls showed an 8.3 degrees increase in pronation of M1 intrinsic torsion (P < .001) and a 4.7 degrees pronated malposition of the first tarsometatarsal joint (P = .02) in HV. A 9.7 degrees supinated malposition of the first naviculocuneiform joint (P < .001) was also observed in HV. Comparing PCFD HV and controls showed a significant increase in pronation of the navicular (respectively, 17.2 ± 5.4 and 12.3 ± 3.4 degrees, P = .007) and a 5.5 degrees increase in pronation of M1 intrinsic torsion (P = .02) in PCFD HV, without malposition of the first tarsometatarsal and naviculocuneiform joints.

Hyperpronation of the M1 head relative to the ground originated from both increases in pronation of M1 intrinsic torsion and first tarsometatarsal joint malposition in HV, although partially counterbalanced by a supinated malposition of the first naviculocuneiform joint. On the other hand, PCFD HV patients showed a generalized pronated position throughout the medial column from the navicular to the M1 head and may be related to the midfoot and hindfoot deformities frequently present in PCFD.

Level III, retrospective comparative study.

Weightbearing computed tomography (WBCT) 3-dimensional measurements may be reliable in assessing hallux valgus (HV). The objective of this study was to compare 2D and 3D WBCT measurements of the M1-M2 intermetatarsal angle (IMA) in patients with HV and in healthy controls. We hypothesized that 2D and 3D IMA measurements would correlate and have similar reliability in both HV and controls.

Retrospective multicenter comparative study included WBCT scans from 83 feet (41 HV, 42 controls). IMA was measured on digitally reconstructed radiographs (DRR-IMA). 3D angle (3D-IMA) and its projection on the weightbearing plane (2D-IMA) were calculated from 3D coordinates of the first and second metatarsals. Intraobserver reliability and intermethod correlations were calculated using intraclass correlation coefficients (ICCs).

Intraobserver reliability was very strong for DRR-IMA (0.95) and 3D-IMA (0.99). Intermethod correlation between the 3 modalities in HV patients ranged from moderate (DRR vs 2D, 0.48; DRR vs 3D, 0.48) to very strong (2D vs 3D, 0.91). Similarly, intermethod correlation in the control group ranged from moderate (DRR vs 2D, 0.56; DRR vs 3D, 0.60) to very strong (2D vs 3D, 0.92).

Measurements for IMA are similar using DRR, 3D and 2D projected angles, with very strong intraobserver reliability and moderate to very strong intermethod correlations. This is the first head-to-head comparison between these measurement modalities in HV. Further investigations are warranted before formulating guidelines for the clinical use of 3D angles.

Level III, case-control study.

Previous studies have demonstrated that preoperative patient-reported outcome measures are associated with postoperative outcomes in foot and ankle surgery, and also in specific procedures such as bunionectomy, flatfoot reconstruction, and total ankle replacement. The primary purpose of this study was to determine if preoperative Patient-Reported Outcomes Measurement Information System (PROMIS) physical function, pain interference, pain intensity, and depression scores were associated with the ability of patients undergoing cheilectomy for hallux rigidus to achieve a minimal clinically important difference (MCID) of improvement.

This retrospective study included preoperative and ≥2-year postoperative PROMIS physical function, pain interference, pain intensity, and depression scores for 125 feet in 118 patients undergoing cheilectomy collected from an institutional review board (IRB)–approved foot and ankle registry database. MCIDs were established using distribution-based methods to determine clinical significance of improvements in PROMIS scores. Receiver operating characteristic (ROC) curves and area under the curve analyses were used to determine which preoperative PROMIS scores were associated with patients meeting the MCID.

ROC analysis found statistically significant areas under the curve (AUCs) for the physical function domain (AUC 0.71), pain intensity (AUC 0.70), and depression (AUC 0.79) PROMIS domains. Subsequent analyses were unable to identify clinically useful 95% sensitivity and specificity preoperative thresholds, with the exception of the 95% sensitivity PROMIS physical function threshold. A preoperative physical function score of greater than 53.2 resulted in a 63% probability of achieving the MCID. The pain interference PROMIS domain did not demonstrate a statistically significant AUC.

Preoperative PROMIS physical function, pain interference, pain intensity, and depression scores are minimally associated with preoperative to 2-year postoperative improvement in patients undergoing cheilectomy. It may be difficult to determine which patients improve from a cheilectomy based on the severity of symptoms they exhibit preoperatively.

Level IV, retrospective case series.

Arthroscopic resection of the talocalcaneal coalition (TCC) has been reported to be associated with good short-term outcomes. However, the midterm outcomes of this approach remain uncertain.

We performed a retrospective study of patients who underwent arthroscopic resection for symptomatic TCC. The patients were divided into 2 groups (group P, patients with isolated posterior facet coalition; and group MP, patients with both middle and posterior facet coalition). The preoperative and postoperative visual analog scale (VAS) scores for pain and American Orthopaedic Foot & Ankle Society (AOFAS) scale scores were calculated. The postoperative AOFAS and VAS scores between the 2 groups were analyzed. Patient satisfaction was also assessed.

Thirty-two patients were included in this study. The mean age at the time of surgery was 26.0±8.5 years, and the mean follow-up period was 56.9±18.0 months. Thirteen (41%) patients were in group P, whereas 19 (59%) patients were in group MP. Postoperative VAS and AOFAS scores improved more significantly than preoperative scores. At the final follow-up, excellent and good subjective outcomes were attained in 26 patients (81%), fair and poor outcomes in 6 patients (19%). There were no statistical differences in the postoperative AOFAS (91.0±7.0 vs 85.8±10.8, P = .532) and VAS score (2.1±1.7 vs 4.0±2.6, P = .537) between patients with the ratio of coalition/posterior facet more than or less than 50%. There were no statistical differences in postoperative VAS score (1.8±1.3 vs 2.6±2.2, P = .236) and AOFAS score (92.5±5.6 vs 89.2±8.7, P = .297) between group P (n=13) and group MP (n=19), either. Three patients (9.4%) had complications, including these notable findings: 1 patient complained of restricted dorsal flexion, 1 with computed tomography–proven coalition recurrence, and 1 with partial tibial nerve injury.

We found that TCC arthroscopic resection was generally associated with reasonable outcomes at midterm follow-up.

Level IV, case series.

Reconstruction of progressive collapsing foot deformity (PCFD) with ankle instability (PCFD class E) remains a substantial challenge to orthopaedic surgeons. We report the outcomes of PCFD class E treated by means of a relatively standard PCFD foot reconstruction approach with the addition of a supramalleolar lateral opening-wedge osteotomy.

We conducted a retrospective study of 13 patients (15 ankles) who underwent this procedure for PCFD class E between 2010 and 2021. Mean follow-up time was 3.8 (range, 2-6.3) years. Clinical assessment was based on the Japanese Society for Surgery of the Foot (JSSF) ankle/hindfoot scale as well as a Self-Administered Foot Evaluation Questionnaire (SAFE-Q). Radiographic assessments, including identification of the mechanical ankle joint axis point and a modified valgus Takakura-Tanaka osteoarthritis grade, were recorded.

The mean JSSF score improved significantly from 45.2 preoperatively to 83.9 postoperatively (P < .001). SAFE-Q scores at the final follow-up were 70.1 for the pain and pain-related subscale, 75.4 for the physical functioning and daily living subscale, 83.0 for the social functioning subscale, 74.4 for the shoe-related subscale, and 78.1 for the general health and well-being subscale. Radiographic measurements showed improvement in the tibiotalar tilt angle (average improvement: 5.4 degrees); orientation of the talus joint line relative to the ground (average improvement: 14.0 degrees); anteroposterior talus–first metatarsal angle (average improvement: 11.2 degrees), talonavicular coverage angle (average improvement: 21.6 degrees), and lateral talus–first metatarsal angle (average improvement: 17.0 degrees). Postoperatively, the mechanical ankle joint axis point moved medially for all patients and into the medial half of the ankle joint for 7 patients. The modified osteoarthritis stage improved postoperatively in most cases.

In this select and relatively small group of patients who had a supramalleolar lateral opening-wedge osteotomy combined with PCFD reconstruction, we measured general improvement in JSSF scores and radiographic variables. Long-term durability of the procedure remains unknown. This procedure may be an option for preserving the ankle joint in treatment of PCFD class E with osteoarthritis of the ankle.

Level IV, therapeutic.

To compare the clinical and radiographic outcomes between the conventional delayed and staged approaches for intra-articular calcaneus fractures in which early definite fixation could not be performed because of severe soft tissue injury.

From January 2015 to May 2019, a total of 32 cases with acute intra-articular calcaneal fractures met criteria and were enrolled in the study. We compared the outcomes of intra-articular calcaneal fractures that underwent delayed internal fixation between groups treated with a conventional delayed approach (non-EF group) vs a temporary medial external fixation (EF group). Clinical outcome measures included a 10-point visual analog scale score, the AOFAS score, and the Foot Function Index. Radiographic outcome measures included Böhler angle, talar declination angle, and calcaneal width. Reduction of the posterior facet was assessed on CT scans.

The first 15 (46.9%) were treated with a conventional delayed approach, and the latter 17 (53.1%) were treated with a staged approach with temporary medial external fixation. Clinical outcomes were not different between the groups at the last follow-up. The time from injury to definite internal fixation was shorter by an average of 3.8 days in the EF group (P = .001). The Böhler angle, talar declination angle, and calcaneal width were not different between the groups before surgery and at the last follow-up. Reduction of the posterior facet on CT scans was significantly better in the EF group than in the non-EF group (good/excellent = 94% vs 60%, respectively, P = .033).

The staged approach using medial external fixation for displaced intra-articular calcaneus fractures could be an effective method to decrease the time to definitive internal fixation and obtain optimal reduction of the posterior facet.

Level III, retrospective case-control study.

The operative treatment of pilon fractures is classically treated with a staged protocol with ankle spanning external fixator, followed by definitive open reduction and internal fixation in order to decrease risk of soft tissue complications and infection. However, treatment of pilon fractures with patrial tibial fixation in addition to ankle spanning external fixation at the time of index procedure may facilitate final fixation while avoiding complications that were associated with acute definitive fixation.

Retrospective cohort series of 113 patients treated for pilon fractures from September 2012 to November 2018 at a single level 1 trauma center. Charts were reviewed to compare patients who underwent traditional management with a staged protocol and those who had a limited tibial reduction and fixation (LTRF) during the index procedure. The main outcome measurement was time to definitive fixation.

Twenty-six percent of patients (29 of 113) had limited tibial reduction and fixation (LTRF) during index surgery. Mean time between index procedure and definitive ORIF was 4.75 days less for LTRF cohort compared to standard stage cohort (10.86 ± 7.44 vs. 15.61 ± 8.59 days, P = .009). The index procedure took on average 51 minutes longer in the LTRF cohort (P < .001), yet definitive procedure operative time was decreased by an average of 98 minutes (P < .001), and overall (index plus definitive) operative duration was shortened by an average of 50 minutes (P = .044). There was no difference in rate of infection between LTRF (3.1%) and traditional treatment (2.5%) (P = .86) or reduction quality (P = .270). There were no nonunions in either treatment group.

Patients who had LTRF had quicker time to definitive ORIF and decreased operative time for definitive ORIF. There was no difference in infection rate, reduction quality, or nonunion rate between groups.

Level IV, Retrospective Cohort Study.

This work used software-guided radiographic measurement to assess the effects of progressive lateral column lengthening (LCL) on restoring alignment in a novel cadaveric model of stage II-B flatfoot deformity.

A stage II-B flatfoot was created in 8 cadaveric specimens by transecting the spring ligament complex, anterior deltoid, and interosseous talocalcaneal and cervical ligaments. Weightbearing computed tomographic (WBCT) scans were performed with specimens under 450 N of compressive load in the intact, flat, and 6-, 8-, and 10-mm lateral column–lengthening conditions. Custom software-guided radiographic measurements of the lateral talo–first metatarsal (Meary) angle, anteroposterior talo–first metatarsal angle, naviculocuneiform overlap, and 2 new measures (plantar fascia [PF] distance and angle) were recorded on digitally reconstructed radiographs. Four anonymized analysts performed measurements twice. Intra- and interobserver agreement was assessed using intraclass correlation coefficients (ICCs).

Six-millimeter LCL restored alignment closest to the intact foot in this new cadaveric model, whereas 10-mm lengthening tended toward overcorrection. The PF line displaced laterally in the flatfoot condition, and LCL restored the PF line to a location beneath the talonavicular joint. Interobserver agreement was excellent for PF distance (ICC = 0.99) and naviculocuboid overlap (ICC = 0.91), good for Meary angle (ICC = 0.81) and PF angle (ICC = 0.69), and acceptable for the talonavicular coverage angle (ICC = 0.65).

In this stage II-B cadaveric flatfoot model, cervical ligament transection was essential to create deformity after the medial hindfoot ligaments were transected. Software-guided radiographic measurement proved reliable; standardized implementation should improve comparability between studies of flatfoot deformity. The novel PF distance performed most consistently (ICC = 0.99) and warrants further study. With this model, we found that a 6-mm LCL restored alignment closest to the intact foot, whereas 10-mm lengthening tended toward overcorrection.

Future joint-sparing flatfoot corrections may consider using a relatively small LCL combined with other bony and/or anatomic ligament/tendon reconstructions.

Patients undergoing total ankle replacement (TAR) often have symptomatic adjacent joint arthritis and deformity. Subtalar arthrodesis can effectively address a degenerative and/or malaligned hindfoot, but there is concern that it places abnormal stresses on the TAR and adjacent joints of the foot, potentially leading to early TAR failure. This study hypothesized that ankle and talonavicular joint kinematics would be altered after subtalar arthrodesis in the setting of TAR.

Thirteen mid-tibia cadaveric specimens with neutral alignment were tested in a robotic gait simulator. To simulate gait, each specimen was secured to a static mounting fixture about a 6-degree of freedom robotic platform, and a force plate moves relative to the stationary specimen based on standardized gait parameters. Specimens were tested sequentially in TAR and TAR with subtalar arthrodesis (TAR-ST_fuse). Kinematics and range of motion of the ankle and talonavicular joint were compared between TAR and TAR-ST_fuse.

There were significant differences in kinematics and range of motion between TAR and TAR-ST_fuse groups. At the ankle joint, TAR-ST_fuse had less internal rotation in early-mid stance (P < .05), with decreased range of motion in the sagittal (–2.7 degrees, P = .008) and axial (–1.8 degrees, P = .002) planes in early stance, and increased range of motion in the coronal plane in middle (+1.2 degrees, P < .001) and late (+2.5 degrees, P = .012) stance. At the talonavicular joint, there were significant differences in axial and coronal kinematics in early and late stance (P < .05). Subtalar arthrodesis resulted in significantly decreased talonavicular range of motion in all planes in early and late stance (P < .003).

In ankles implanted with the TAR design used in this study, kinematics of the ankle and talonavicular joint were found to be altered after subtalar arthrodesis. Aberrant motion may reflect altered contact mechanics at the prosthesis and increased stress at the bone-implant interface, and affect the progression of adjacent joint arthritis in the talonavicular joint.

These findings may provide a correlate to clinical studies that have cited hindfoot arthrodesis as a risk factor for TAR failure.

Detection of Lisfranc malalignment leading to the instability of the joint, particularly in subtle cases, has been a concern for foot and ankle care providers. X-ray radiographs are the mainstay in the diagnosis of these injuries; thus, improving the performance of clinicians in interpreting radiographs can noticeably affect the quality of health care in these patients. Here we assessed the performance of deep learning algorithms on weightbearing radiographs for detection of Lisfranc joint malalignment in patients with Lisfranc instability.

In a retrospective study, 640 patients with Lisfranc malalignment leading to instability were recruited plus 640 individuals with uninjured feet and healthy Lisfranc joint as the control group. All radiographs were screened by orthopaedic surgeons. Two deep learning models were trained, validated, and tested (in a ratio 80:10:10) using a single-view (anteroposterior) and 3-view (anteroposterior, lateral, oblique) radiographs. The performances of the models were reported as sensitivity, specificity, positive and negative predictive values, accuracy, F score, and area under the curve (AUC).

No significant differences were observed between the patients and the controls regarding age, gender, race, and body mass index. The best deep learning algorithm outperformed our human interpreters (<1% vs ~10% misdiagnosis), 94.8% sensitivity, 96.9% specificity, 98.6% accuracy, 95.8% F score, and 99.4% AUC.

Deep learning methods have shown promising potential in acting as an assistant interpreter of radiographic images in patients with Lisfranc malalignment. Developing these algorithms can hasten and improve the accuracy of diagnosis and reduce further costs and burdens on the patients and health care system.

Level III, case-control Machine Learning study.