Mobile-bearing total knee arthroplasty was developed to provide low contact stress and reasonably unrestricted joint motion. We studied the results of a cementless, posterior cruciate ligament (PCL)-retaining total knee arthroplasty (TKA), with a mobile-bearing insert in rotation and anterior-posterior (AP) translation (Innex^® Anterior-Posterior Glide, Zimmer).

Kinematic analyses were performed on a series of 51 primary TKA. The patients’ mean age was 71±8 years at operation. Patients were studied at 23 months average follow-up with weight-bearing radiographs at full-extension, 30° flexion and maximum flexion (“lunge” position). Three dimensional position and orientation of the mobile-bearing relative to the femoral and the tibial component during flexion were determined using model-based shapematching techniques.

The average weight-bearing range of implant motion was 110°±14°. In flexion, the mobile-bearing was internally rotated 3°±3° with respect to the femoral component (p< 0.0001) and the tibial tray was internally rotated 5°±7° with respect to the mobile-bearing (p< 0.0001). On average, the mobile-bearing did not translate relative to the tibial base plate from full extension to 45° flexion [0±2 mm (range −5 mm to 6 mm)]. However, the mobilebearing did translate anteriorly 1±2 mm (range −2 mm to 9 mm, p< 0.0001) between 45° flexion and maximal flexion.

We conclude that the mobile-bearing insert showed a progressive increase in internal rotation during flexion. Most of this rotational mobility occurred between the mobile insert and the tibial base plate. With flexion, AP translation did occur between the femoral component and mobile-bearing, and between the mobile-bearing and tibial base plate, but mobile-bearing translation was unpredictable with this unconstrained design.

Purpose of the study: The correct position of the femoral and tibial tunnes for anterior cruciate ligament (ACL) reconstruction is a determining factor for favorable outcome. We used a novel computer-assisted system which enables intraoperative localization of the tunnel centers on the lateral view of the knee before drilling. This technique uses fluoroscopy combined with a passive system for computer-assisted image acquisition and processing to provide the surgeon with the desired positions. We report the anatomic and clinical results observed in a prospective series comparing this technique with the classical technique of independent blind tunnels.

Material and methods: Thirty-seven patients underwent computer-assisted surgery and 36 classical surgery without computer assistance performed by a senior surgeon. Mean patient age was 27 years in both groups. The patients were reviewed at mean 2.2 years (range 1–4.5 years). Data recorded included the KT-1000 laxity, radiographic drawer and the IKDC score (1999).

Results: Mean time from ACL tear to reconstruction was 30 months in both groups. Computer assistance increased operative time 9.3 minutes (range 4–13). The IKDC score was 67.9% A, 29.7% B, and 2.7% C in computer-assistance surgery patients and 60% A, 37.1% B and 2.9% B for classical surgery patients. The mean IKDC function score was 89.7/100 for the computer-assisted patients and 89.5/100 for the others. Mean manual maximal laxity (KT-1000) was 7 mm before surgery and less than 2 mm at last follow-up. Differential laxity was less than 2 mm in all patients who underwent computer-assisted surgery and in 97.7% of the others. The mean differential laxity for the medial compartment as measured on the postoperative stress films was 2.4 mm (range 0–12 mm) for computer-assisted surgery patients and 3 mm (range 0–10 mm) for the others. In the computer-assised surgery patients, the femoral tunnels were centered on a smaller area. There was not significant difference in the IKDC score, the KT-1000 findings and the stress x-rays between the two techniques.

Conclusion: The results of these two techniques in this report are similar to data reported in the literature. Computer-assistance enables more accurate and reproducible tunnel positioning with no significant clinical impact.

Purpose of the study: The aim of this retrospective study was to analyze the results obtained with the femoropa-tellar self-centering prosthesis and to confirm its usefulness and limitations for the treatment of femoropatellar osteoarthritis.

Material and methods: This continuous series of 57 patients, mean age 65.6 years underwent surgery fro 1986 to 2003 for implantation of a self-centering prosthesis. Most (60%) presented osteoarthritis on a dysplasic knee; others presented primary osteoarthritis (31%) or trauma sequelae (9%). Mean follow-up was 74.5 months (range 12–180 months). Functional outcome was assessed with the IKS and activity daily living (ADL) scales. The radiographic study enabled 3D analysis of the implant position.

Results: At last follow-up, the mean IKS score was 157.3/200. The mean ADL score was 73.8 (range 48.8–96.3). The IKS score gave 66.7% good and the ADL score, which takes into consideration all daily life functions, 57% very good outcome. Outcome was best in patients with trochlear dysplasia. The eleven knees requiring revision for a total knee arthroplasty (24%) had initially a narrower trochlear angle (p=0.02) and a thicker patella (p=0.02). In these patients, the initial prosthesis protruded more anteriorly (p=0.004) and the horizontal axis was greater (p=0.02).

Discussion: The results of this series are less satisfactory than in the literature, but assessments may depend on the scores used since there was a 10% difference between the ADL and IKS scores. The results were better in the group of patients undergoing surgery for osteoarthritis due to dysplasia, in agreement with De Cloedt and Argenson. Analysis of the radiological parameters enabled identification of technical errors leading to failure. A successful femoropatellar prosthesis depends basically on two factors: technical precision and patient selection.

Conclusion: These results led us to avoid widening indications for femoropatellar prostheses, which are already in our experience rather limited (2%). For us, patients aged 50–70 years with advanced-stage femo-ropatellar osteoarthritis due to dysplasia and no other anomaly and who have not responded to conservative treatment would be the population of choice. The lack of any technical problem for revision total knee arthroplasty is an argument in favor of the femoropatellar prosthesis.

Purpose of the study: The pathophysiology of reflex dystrophy or type I complex regional pain syndrome remains poorly understood, but the cost is considerable in terms of public health expenditures both for programmed and emergency orthopedic surgery. We present a historical cohort assessed to evaluate the usefulness of vitamin C for the prevention of reflex dystrophy in programmed foot and ankle surgery.

Material and methods: The study included two groups of patients treated in two successive periods: July 2002 to June 2003 and July 2003 to June 2004. All patients underwent foot and ankle surgery performed by the same senior surgeon. Diabetic feet were excluded. The first group (185 feet, 177 patients) was not given any particular preventive treatment. The second group (235 feet, 215 patients) was given one gram vitamin C for 45 days. The diagnosis of reflex dystrophy was retained on the basis of clinical and radiological arguments noted at follow-up visits with the operator. Several factors were studied: gender, age, type of disease condition, history of reflex dystrophy, psychological context, duration of tourniquet, cast immobilization.

Results: Reflex dystrophy occurred in 18 feet in group

1 (9.6%) and in 4 (1.7%) in group 2. The difference was significant. Presence of a history of dystrophy was significantly associated with development of dystrophy (RR=10.4). A psychological context appeared to increase the risk of dystrophy (RR 2.6) but did not reach significance. There was no statistical relationship with age, gender, duration of tourniquet, type of disease condition, or surgical procedure performed.

Discussion: Vitamin C has been found to be effective in the prevention of reflex dystrophy after wrist fractures. Data in the literature is scarce on dystrophy of the foot and ankle. Our study provided objective evidence of the usefulness of vitamin C for the prevention of reflex dystrophy in foot and ankle surgery patients, a complication frequently observed in our control group (9.6%). The psychological context and history of dystrophy increase the risk of dystrophy.

Conclusion: Vitamin C is associated with a lower risk of reflex dystrophy in the postoperative period after foot and ankle surgery. We advocate preventive treatment with vitamin C.

Purpose of the study: Reconstruction of the anterior cruciate ligament (ACL) has become a common procedure. We compared two randomized series: intra-articular (Kenneth-Jones) versus intra- and extra-articular (MacInJones).

Material and methods: From January 1995 through March 1998, 73 knees were treated surgically for differential medial laxity measured at 7 to 12 mm on passive stress x-rays in 20° flexion. Group 1 (ACL reconstruction alone) included 34 patients (aged 27.1±7.5 years). Group 2 (ACL reconstruction plus extra-articular plasty) included 29 patients (aged 28.5±12 years). Function was scored 72% in group 1 and 68% in group 2 at mean seven years follow-up (102 and 93 months follow-up respectively). Anterior laxity was measured radiographically and with KT-1000 and the position of the tunnels was assessed according to Aglietti.

Results: According to the IKDC, functional outcome was 83.9±3.1 in group 1 and 83.3±3.6 in group 2. The overall IKDC classification was 0A, 57.8% B, 26.3% C, and 15.7% D for group 1 and 58% A, 52.9% B, 29.4% C, and 11.7% D for group 2. The pivot-shift test was negative in 61.1% of group 1 knees (27.7% grade 1 and 11.1% grade 2) and negative in 83.3% of group 2 knees (16.6% grade 1). In group 1, the radiological drawer showed 46.09% improvement in the differential laxity for the medial compartment and 41% for lateral compartment. In group 2 the corresponding improvements were 44.8% and 44.6%. There was no difference in tunnel position between the two groups.

Discussion: The two-year results of this series did not provide any evidence favoring a clear advantage of complementary lateral plasty. At seven years follow-up, the pivot-shift test appeared to favor associated lateral plasty (p=0.09), but with no significant difference in laxity for the two compartments.

Conclusion: Anterior laxity was only incompletely controlled by both reconstruction techniques. In this context of relatively limited laxity (7–12 mm initially), at seven years follow-up there was no certain advantage of complementary lateral extra-articular plasty in combination with ACL reconstruction.

Purpose: Failed anterior cruciate ligament reconstruction as defined by recurrent pathologic laxity, is increasingly commonplace. We reviewed 77 patients who had undergone unsuccesful anterior cruciate ligament surgery to correct persisting instability, and who underwent revision surgery.

Material and Methods: During the first operative treatment, were used synthetic ligament in 18 cases, autograft in 54 cases, extra-articular plasty in 4 cases, allograft in 1 case and primary repair in 1 case. For revision, we used autograft in all cases according to differents anatomicals factors: 41 patellar tendons, 15 quadriceps tendons and 17 hamstring tendons. 46 patients had meniscectomy during one of the two surgeries ; 19 patients had cartilage lesions (grade 3 or 4).

For clinical evaluation, we used the IKDC score (1999), and laxity measurement with the KT-1000 arthrometer and stress X-rays.

The mean follow-up was 24 months.

Results The mean IKDC subjective score was 71,5 and 75% of knee were considered as normal or nearly normal. The surgery was successfull in objectively improving the stability in most of patients with a KT-1000 differential maxi-manual of 2 1,7 mm.

We found no statistical difference between the three groups of graft used for revision. The results are a trend toward less good results, when patients had a meniscec-tomy. Subjectively the result were worse in cases of cartilage lesion. In fact, no patient who had grade IV lesion returned to there previous level activity (pre-operative level activity). The worse results are in the group of failed synthetic ligaments.

Conclusion ACL revision surgery leads to poorer results than primary surgery.

There was no clinical difference for the revision, whether we used autograft of patellar tendon, quadriceps tendon or hamstring tendon with an adapted fixation device.

On the other hand, meniscal or cartilage lesion or the use of synthetic grafts are factors of poor clinical outcome.

Purpose: The procedure described in 1986 in the Revue de Chirurgie Orthopédique was used for 24 knees operated on since 1982.

Operative technique: A single anterior incision measuring 18 cm and a 28-cm transplant using the patellear tendon, a slit of patellar bone, and the quadriceps tendon. The patellar bloc is fixed in the tunnel bored under the spinous process, using the patellar tendon to reconstruct the posterior cruciate ligament (PCL) and the quadriceps tendon to reconstruct the anterior cruciate ligament (ACL). It crosses the lateral condyle and prolonged to the Gerdy. The bone is blocked but can be fixed with a screw. Since the initial description, a second strand for a lateral plasty has been added from the condyle towards the posterior tibia (six cases). Two strands can be used for the PCL. The tunnel can run oblique upwardly to facilitate passage of the transplant behind the tibia (a 30 cm transplant is needed). Operative time 177±39 minutes.

Material and methods: The series included 24 knees (75% men) (13 right and 3 bilateral procedures). The patients were aged 23±6 years at the time of the accident which was a sports accident in seven and an occupational accident in one. Time from the accident to surgery was 35±41 months. Prior surgery had been performed on fourteen knees (five sutures, five artificial ligaments, two ACL reconstructions). Three patients had fibular nerve palsy. Radiologically, the anterior drawer of the medial and lateral compartments was 124 and 185 mm; the posterior drawer was 174 and 126 mm. Lateral laxity predominated in 17 knees, medial laxity in seven and mixed laxity in eleven. The frontal axe was normal in five knees. Associated procedures depended on the laxity: three procedures to tighten the lateral popliteal ligament, three to tighten the posterior capsule of the femur, and seven to tighten the medial capsule of the femur. There was one meniscal suture and four regularisations. The knee was mobilised postoperatively and weight-bearing with assistance was allowed except in five cases with a weak reconstruction or associated osteotomy.

Results: Mean follow-up was 4.5 years (1–20). Flexion was 130±11° and function was acceptable except for three patients who were able to resume their occupational activities but little sports activity. None of the patients was disappointed. Mean gain in anterior drawer was 5 mm for both compartments; gain for the posterior drawer was 6 mm medially and 3 mm laterally. Peripheral laxity was corrected best.

Discussion: There are no data in the literature on cruciate reconstruction in patients with chronic laxity. This series is small but does have 20 years follow-up and provided precious information for improving results. It would be particularly interesting to use two strands for the PCL reconstruction, to use more peripheral procedures, especially with two lateral plasties and medial plasties, and to use osteotomy.

Conclusion: Cases of double cruciate tears seen late are exceptional and appropriate treatment has not been standardised. Instead of using to distinct incisions with two transplants for the two cruciates, as would be possible with the classical techniques, we propose a single incision and a single transplant with peripheral plasties depending on the radiographic laxity.

Purpose: The posterior cruciate ligament (PCL) has two strands, an anterolateral strand (AL) and a posteromedial strand (PM). Its femoral insertion fans out over 3 cm and cannot be replaced by a unique transplant during surgical reconstruction. The purpose of this study was to define the anatomic centre of the femoral insertion of each stand in order to identify precise and reproducible landmarks for the bone tunnels (one for each strand) used to fix the transplants during reconstruction of the PCL.

Material and methods: A metallic landmark was placed on the centre of the femoral insertion of the two PCL strands in ten cadaver knees. The Metros software package was used to analyse the digitalised radiograms of each knee to determine the position of the strands on the medial condyle. Intra- and inter-observer variability was determined.

Results: The AL strand was situated 31.6 + 2.45% (47.2 + 6.02% for the PM strand) from the anterior border of the notch or 41.18 + 2.73% (54.46 + 5.07% for the PM) from the anterior border of the medial condyle relative to the Blumensaat line and 16.12 + 4.45% (33.68 + 7.2 for the PM) from the apex of the notch.

Discussion: Clinical and objective results of reconstruction depend on the ideal, basically femoral, position of the PCL insertions. Intraoperative identification of the ideal point for the femoral insertion can be improved with measurements made on cadaver knees. The values observed in the present study are reproducible. Presented in the form of percentages of length limiting the errors related to patient morphotype can be integrated into navigation systems.

Purpose: Several studies have reported results of anterior cruciate ligament (ACL) reconstruction showing the influence of the position of the transplant. We choose the anatomic position. In order to optimise this position, we developed a navigation system using intraoperative fluoroscopic imaging.

Material and methods: Thirty-five patients underwent ACL reconstruction using the computer-assisted technique. We used a fluoroscope connected to the computer equipped with an acquisition module and an image processing module which captures the lateral view of the knee provided by the fluoroscope. Surgery was performed arthroscopically. After inserting landmarks (mini-screws for the femoral end and pins for the tibial end) on the theoretically ideal positions, the computer determined the theoretical anatomic position of the tibial and femoral insertions of the ACL. The position was then validated or modified and re-validated. Postoperative radiograms (lateral view of the knee) were used to analyse the position in relation to the anatomic centre of the ACL insertion. Two groups of patients matched for sex and age were compared, 35 patients undergoing ACL reconstruction with the navigation technique versus 35 patients undergoing the same procedure without navigation. Results were compared with the nonparametric Wilcoxon test.

Results: Computer-assisted positioning provided a mean difference of 5.1±1.3 mm between the centre of the transplant and the theoretical ideal point compared with 7.7±1.9 mm without assistance.

Discussion: Comparison using adapted statistical tests (Wilcoxon text) demonstrated a significant difference (p=0.001) between the two groups. These findings demonstrate the improved precision and reproducibility achieved with the navigation technique.

Conclusion: This technique allows the surgeon to obtain an excellent precision of the transplant insertion with excellent reproducibility. There are two drawbacks: the longer operative time (mean 15±7 minutes) and radiation exposure. But this technique is simple, easy to use and low-cost.

Aims: Functional results of anterior cruciate ligament (ACL) is closely related to the anatomical position of the graft. Femoral or tibial miss placements are associated with increase laxity, decrease range of motion. Based on anatomical dissections we developed a triangle method able to be used as fluoroscopic intra-operative landmark to increase the reproductibility of the placements. Methods: A lateral X ray of normal knee is mandatory before the surgical procedure. During the ACL reconstruction using patellar tendon a pin is placed to determine the tibial tunnel, and a small screw is inserted at the femoral ACL location. Under imagine intensifier, a lateral X ray of the operated knee is realised. The picture is exported and analysed on a PC computer. The surgical placement is compared with the ACL center position according to the triangle method. 35 knees have been operated and compared with 35-paired knees operated with the same technique and operator. Results: The mean distance between the ACL center and the surgically chosen femoral position is 7.5 ± 1.9 mm for the standard surgery and 4.9 ± 1.3 mm for the fluoroscopy based surgical navigation. The Wilcoxon test for small-paired series indicates a statistical significance (p=0.001). The mean extra operative time is 15 ± 7mn. Conclusion: This technique is simple, easy and rather fast. It gives to the surgery a significative improvement for positioning the ACL graft on the femoral side. It has to be confirmed for larger and by long term clinical results.

Aims:ACL revision surgery is a second-generation type of knee ligament surgery. The artificial ligaments, the imperfect mastering of arthoscopic assisted surgery and the absence of clinical and radiological analysis of peripheral laxities are among the main factors for failures. Methods:43 patients of a mean age of 29y were previously operated between 1 and 5 times.14 artificial ligaments, 23 patellar tendons and 4 hamstrings tendons failed as a first ACL reconstruction. The mean time between the first operation and the index revision was 44 months. In 6 cases an additional HTO was used. Different tendon grafts were used: quadriceps 11, patellar 19 and hamstrings 8. In 3 occasions an additional extra articular reefing was used. The patients were reviewed by one independent observer using KT 1000, Stress X rays, IKDC form (2000). The mean follow up is 35 months (11–123)Results:The IKDC score in pre operative time was 19 D, 21 C and 1B. At the review the score is 2A, 28 B, 9C, 2D. The functional IKDC form show 37.5% of remaining pain, 44% of stiffness sensation and 12.5% of instability. The mean functional improvement is 44%. The mean laxity improvement is 5.3mm for KT1000 and 4.5mm for stress X rays. Conclusions:Revision ACL surgery is not as good as primary surgery. The reconstruction is technically difficult and must be ‘à la carte’ in order to take in account several simultaneous problems: bone defect, cartilage abnormalities, skin and ligament insufficiencies.

Purpose: Constitutional factors responsible for hallux falgus and hallux rigidus remain unclear. The purpose of this work was to compare the radiological feature of the forefoot in three populations with “normal” feet, hallux rigidis, and hallux valgus.

Material and methods: One standard protocol was used within the same unit to obtain dorsoplantar AP views of the foot in the standing position in all subjects. Fifty “normal” feet, with no apparent deformation, callosity, or pain, were selected among the orthopaedic unit personnel; mean age of the 25 subjects was 30.3±9.6 years, and 44% were women. The 30 patients with hallux rigidus were operated on at a mean age of 57.4±10.7 years, and 48.4% were women. The 50 patients with hallux valgus were operated on at a mean age of 50.8±12.8 year and 92% were women.

All radiograms were digitalised (Vidar VXR-12 plus) and analysed by four observers using the FootLog software which provides semiautomatic measurements. The following parameters were recorded: distance between the lateral sesamoid and the second metatarsal (LS-M2), the M1P1 angle (for the diaphyseal and mechanical axes of M1), the diaphyseal and mechanical distal metatarsal articular angle (DMAA) of M1, Meschan’s angle (M1–M2–M5), the distance between a line perpendicular to the axis of the foot drawn through the centre of the lateral sesamoid and the centre of the head of M4 (MS4–M4) (a corrective factor was introduced for the MS4–M4 distance to account for the displacement of the lateral sesamoid in hallux valgus), the M1 index = d1-D2 (length of the head of M1/MS4 – length of the head of M2/MS4), maestro 1 = d2–d3, maestro 2 = d3–d4, maestro 3 = d4–d5. The measured parameters were recorded automatically on an Excel data sheet and statistical analysis was performed with SPSS 9.0.

Results and discussion: Intra- and inter-observer reproducibility of measurements and morphological classifications were excellent. The LS/M2 distance was comparable in the three populations, proving that the lateral sesamoid is relatively fixed compared with the M2 and enabling its use as reference for the MS4 line. The Meschan angle did not discriminate between the three populations, likewise for the mean M1/M2 index, the M1P1, M1M2, and DMAA angles which were different in the three populations; there were 2° to 3° variations for the mechanical or shaft axis. The morphotype analysis demonstrated objective evidence of morphological differences of the forefoot in the three populations. The hallux rigidus group showed a predominance of the index plus and plus-minus with long M23 lateral patterns, while the hallux valgus group exhibited a predominance of M4M5 hypoplasia.

Discussion: Morphotypic definition of the metatarsals is an interesting approach providing a measurable way of interpreting forefoot disorders and guide surgical correction. These results should be confirmed with measures in larger series, which can be accomplished with FootLog software. It would also be useful to combine radiological studies with baropodometric studies.

Purpose: We analysed knee prostheses preserving the posterior cruciate ligament (or both cruciates) performed after tibial osteotomy. We determined outcome as a function of any tibial callus deformation created by the tibial osteotomy.

Material and methods: This retrospective study included a consecutive series of 56 knee prostheses with preservation of the posterior cruciate ligament (n=43) or both cruciate ligaments (n=13). The patients had undergone prior tibial osteotomy for valgisation (n=47) or varisation (n=9). Seven groups were defined as a function of the preoperative tibial angle prior to TKA. The angle were measured with telegonometry. Minimum follow-up was one year, mean follow-up 4.1±2.8 years.

Results: The tibial tuberosity was raised in 15 cases. If there was major valgus or rotation deformation, tibial osteotomy was associated with the prothesis (n=9). At last follow-up, the mean IKSg, IKSf and HSS scores were 81.5, 77.6, and 82.3 respectively. The mean femorotibial angle was 177.4±4.2°. The mean tibial angle was 87.8±3° and the mean femoral angle was 89.8±2°. Preoperative tibial deformation was not influenced by clinical results. In case of preoperative tibial deformation situated between 5° valgus and 5° varus, operation time, blood loss, and femoraotibial axis at last follow-up were not significantly different. To correct for tibial valgus greater than 7°, tibial osteotomy was associated with prosthesis implantation during the same operative time in six out of thirteen cases. For preoperative tibial varus greater than 5°, the femorotibial axis was less well corrected.

Discussion: These clinical results were comparable to those reported in other series with preservation or not of the posterior cruciate ligament. Correction of the femorotibial angle was less satisfactory than in certain series, but the deformation and the surgical history were among the most marked in the literature. Preservation of the posterior cruciate ligament (or both cruciates) appears to have increased the technical difficulties for upper tibia exposure and position of the tibia implant. For tibial callus with valgus greater than 7°, the prostheses cannot be expected to provide a solution alone and osteotomy should be associated. For tibial callus with 5° or more varus, the indication for associated tibial osteotomy merits discussion.

Purpose: Several techniques have been described for the treatment of unstable fractures of the upper humerus. None appear to be appropriate for subtuberosity fractures, associated or not with a fracture of the trochiter or impacted valgus cephalotuberosity fractures, allowing a stable fixation with anatomic reduction while preserving blood supply to the bone fragments. For the last year, we have studied prospectively a percutaneous minimally invasive technique for this type of fracture. Our preliminary results are analysed here.

Material and methods: We used this technique for 12 patients aged 30 to 87 years with five displaced subtuberosity fractures, six subtuberosity fractures with a trochiter fracture and one impacted valgus cephalotuberosity fracture. Excepting the cephalotuberosity fracture, the joint fragment of the head was reduced by external manipulation under image amplifier guidance. Fixation was achieved with two 25/10 threaded pins inserted percutaneously in retrograde fashio from the anterolateral cortical to the humeral shaft. For eight cases, a third pin was inserted percutaneously from the trochiter to the medial cortical of the humerus. When percutaneous reduction of the trochiter was impossible (three cases) and for the cephalotuberosity fracture, we used a minimal transdeltoid lateral incision to reduce the trochiter and achieve reduction.

Results: Reduction was very satisfactory in all cases. There was one superficial infection that required pin withdrawal at three weeks leading to the only secondary displacement that was minimal and tolerable. We removed the pins after a mean two months. There were no nonunions. Seven patients have a follow-up greater than six months and exhibited a Constant score of 87% (71% to 100%).

Discussion: This techniques has provided very satisfactory results for rapid and stable fixation of the cephalic fragment without loss of blood supply and with a material easy to remove.

Conclusion: These early results are very encouraging and incite us to pursue this technique and analyse long-term results.

Purpose: Plantar pressure sores can lead to metatarsalgia depending on the patient’s activity level and age and on the status of the muscle-tendon system and the morphology of the forefoot. In 1995, Tanaka and Maestro attempted to quantify the relative lengths of the metatarsals. The purpose of this work was to check the results reported by Maestro and to try to define a morphotype classification of the metatarsals.

Material and methods: We analysed two series of normal feet: no apparent deformation, no callosity, no pain, no history of trauma or surgery. Fifty “normal” feet were selected among the personnel of the orthopaedics unit. Mean age of the 25 subjects was 30.3 ± 9.6 years, 44% were women. This series was compared with 34 “normal” feet reported by Maestro (age 55.2 ± 17.2 years, 62% women) used to define criteria for geometric progression (1995). A standing dorso-plantar radiograph was obtained with the same protocol for all patients. All radiographs were digitalized with a Vidar VXR-12 plus, then analysed by two observers with the semi-automatic FootLog measurements. The following measurements were recorded: SM4-M4 (distance between the line passing through the centre of the lateral sesmoid and perpendicular to the foot axis and the centre of the M4 head), M1 = d1 – d2 (length of the M1/SM4 head – length of the M2/SM4 head), Maestro criteria 1 = d2 – d3, Maestro 2 = d3 – d4, and Maestro 3 = d4 – d5.

Results: An SM4 line passing through the mid third of the M4 head (+2mm proximally / centre M4 head / −4 mm distally) as normal. The notion of row 2 geometric progression was conserved by tolerating 20% variation (Maestro 1 ± 1 mm, Maestro 2 ± 1mm, Maestro 3 ± 2 mm). Feet were classed in four metatarsal morphology types with subgroups: normal feet (line SM4 passing through the mid third of the M4 head – geometric progression) – long M23 (SM4 line centred on the mid third of M4 – but alteration of the geometric progression) with four subgroups (long M2, long M3, long M2-3, long M23 long 2) – M4M5 hypoplasia (distal SM4 line / at mid third of M4) with four subgroups (by geometric progression: long M2, long M23, long M23 long M2) – others (long M1: M1 > 3.3 mm causing distalization of SM4).

Discussion, conclusion: FootLog enables rapid radiographic measurements with excellent precision and intraobserver (variations from 0.1 to 0.2 mm and 0.1 to 0.5°) and interobserver (variations from 0.1 to 0.5 mm and 0.1 to 1°) reproducibility. In the two series of clinically “normal” feet, the measured parameters were strictly comparable. Radiologically, 31% were “normal”, and the others (30% long M23 – 37% M4M5 hypoplasia – 2% others) could be considered as predisposed to potential forefoot disorders. Finally only 48% of the subjects had the same morphotype for both feet. This study adds further precision to earlier qualitative evaluations of the forefoot architecture.