In the current austere financial climate within the NHS where local healthcare Trusts are reimbursed in a Payment by Results system it is important that we accurately identify the costs associated with surgical procedures.

We retrospectively reviewed data of 589 consecutive patients undergoing lower limb arthroplasty surgery and recorded their age, BMI and co-morbidities. The effect of these parameters on operative duration and length of stay (LOS) was analysed.

We demonstrate that for a 1 point increase in BMI we expect LOS to increase by a factor of 2.9% (p<0.0001) and mean theatre time to increase by 1.46 minutes (p<0.0001). We also show that for a l-year increase in age, we expect LOS to increase by a factor of 1.2% (p<0.0001).

We have calculated the extra financial costs associated with this and believe that the current OPCS coding system for obesity underestimates the financial impact of increasing BMI and age on lower limb arthroplasty Trusts are being inadequately reimbursed.

The results of this study have been used to produce a chart that allows prediction of LOS following lower limb arthroplasty based on BMI and age. We also believe that the data produced is of use in planning operating lists.

Recent NICE guidelines have suggested abandoning the Thompson hemiarthroplasty (TH) in favour of a ‘proven prosthesis’ such as the Exeter Trauma Stem (ETS). This is controversial because of significant cost implications and limited research assessing outcomes of the ETS. The aim of this study was to assess the treatment of intracapsular neck of femur fractures with the TH.

Between 2002 and 2006 (minimum 5 year follow-up), 431 cemented TH's were performed. Death rate at 1 year and 5 years were 26.0% and 67.7% respectively. Dislocation (1.4%) and infection (0.2%) rates were low and revision rate was 1.2%. Comparison was made to Bipolar hemiarthroplasties over the same period (total 194). These had lower rates of dislocation (0.5%) and infection (0.5%) with a significantly higher (3.6%) revision rate.

We feel that the TH remains the current gold standard treatment for intracapsular fractures, in appropriate patients, due to low complication and revision rates. Modern implants may provide better function or longevity, but there is no evidence in the literature to support abandoning the TH. Surgeons should assess patients and decide on its use, despite NICE guidelines, as it remains a cost effective treatment method, particularly for older, less mobile and cognitively impaired patients.

We designed this study to determine the clinical evidence to support use of the five degree tibial extra-medullary cutting block over the zero degree cutting block.

We identified three groups of patients from the databases and clinical notes at St Michaels Hospital, Toronto. Group one were primary total knees performed using the five degree cutting block, group two were primary total knees performed using the zero degree cutting block and the third group were computer navigated primary total knees. Patients in all three groups were age and sex matched. The senior author advocating use of the five degree block aimed to obtain a five degree posterior slope. The senior author who advocated the use of computer navigation, or the traditional zero degree cutting block, aimed to obtain a three degree posterior slope. All operations were performed by residents or clinical fellows, under the supervision of the senior authors. Patient radiographs were assessed to obtain the optimal direct lateral view obtained and they were saved on a database. Two independent blinded researchers assessed the posterior slope using Siemens Magicweb Software Version VA42C_0206. Two methods were used and the results averaged. The average posterior slope for the navigated total knee replacements was 0.1 degrees (−2 to 4). The average posterior slope for the five degree cutting block was 5.2 degrees (−2 to 16). The average posterior slope for the zero degree block was 3.79 degrees (−2 to 13). Computer navigated knee arthroplasty patients had significantly less variation in outlier measurements compared to the traditionally jigged arthroplasty patients. They were however, less accurate. The five degree cutting block tended to provide a more consistent posterior slope angle, but both the five degree and zero degree cutting blocks had variability in outliers. Computer Navigated Total Knee replacement provides a more consistent and reproducible tibial cut with less variability in alignment than extra-medullary jigs. The traditional five degree cutting block tended to provide a more reliable five degree posterior slope than the zero degree block, but was still subject to outliers.

The aim of this study was to determine the mid-term survival and functional outcomes of the Scorpio Total Stabilised Revision Knee prosthesis.

Sixty seven prostheses were implanted between November 2001 and April 2008. 42 females and 23 males. Average patient age was 67.9 (37-89). Outcomes were assessed with WOMAC (Western Ontario and McMaster Universities Osteoarthritis index), Knee Society Scores, Short Form-8 scores, patient satisfaction and radiological review. Average follow-up was over 3 years (8-93mths) with 95% follow-up.

One patient died post operatively and 4 patients from 18 months to 5 years post-operatively. Average body mass index was 32.9 (21.5- 55.1). 65% (42 patients) of patients operated on had a Body Mass Index of greater than 30. 48 patients were ASA 3 or greater.

Thirteen second stage revision arthroplasties were performed after treatment for infected arthroplasty surgery. Twenty six prostheses were revised for aseptic loosening. Eight prostheses were revised for stiffness and 9 for worn polyethylene inserts. Five prostheses were revised for symptomatic tibio-femoral instability/ dislocation and one for patello-femoral instability. Two revisions were performed for peri-prosthetic fractures and 2 for previously operated tibial plateau fractures.

Seven patients required tibial tubercle osteotomy and seven a rectus snip. Thirty one patients had greater than a 15mm polyethylene insert. The average KSS increased from 49 pre-operatively to 64 at 7.5 years. The average KS function score increased from 21 to 45. 68% (44) of patients had other significant joint involvement which affected daily function. 24% of patients were unsatisfied with the outcome. 89.5% of patients radiographs were assessed for loosening or subsidence. 51% of femoral components and 36% of tibial components had radiosclerotic lines. The surface area of each implant including the stem was measured on antero-posterior and lateral images. The degree of lucency was calculated as a percentage and in mm from the component.

Two prostheses (3%) were revised for deep infection, one (1.5%) for stiffness and one for aseptic loosening (1.5%). Complications included a popliteal artery injury, two superficial wound infections, and one patella tendon avulsion.

Survival rate for revision of prosthesis was 87% at 7.5 years and 90% excluding infection. Success of second stage revision arthroplasty after treatment of infection was 92%.

Introduction

Paediatric hip fracture accounts for less than one percent of paediatric fractures. Previous studies report complication rates between 20 and 92%.

The purpose of this study was to evaluate 3 methods used to produce posterior tibial slope.

We reported the first single surgeon series comparing outcome of microscopic and open primary single level unilateral lumbar decompression or discectomy. We aimed to determine any difference in outcomes between the two techniques.

Forty-six decompressions were performed with use of an operating microscope (microscopic), and forty without (open) at two different hospitals. All procedures were performed by the senior author. Information was obtained by analysis of the patients' notes. The average age of the patients in both groups was comparable. Operating time was shorter in the microscopic group (68min, range 30-130) compared to the open group (83 mins, range 30-180). Dural tear rate was 4.3% with use of a microscope (0% symptomatic dural tear rate) and 7.5% without (2.5% symptomatic dural tear rate). Nerve damage incidence was 0% with use of a microscope and 5% (two patients) without. One of these was a neurapraxia and the patient made a full recovery. Wound infection rates, diagnosed on grounds of clinical suspicion, were 4.3% and 2.5% for microscopic and open respectively. There were no incidences of deep infection or post-operative discitis. Average inpatient stay was under 48 hours in both groups.

Using the modified Macnab criteria, results using the microscope were 0% poor, 14% fair, 32% good, and 55% excellent. The results for the open group were 0% poor, 10% fair, 37% good and 53% excellent. Average follow-up was six months (1-19) for the microscope group, and seven months (2-16) for the open group.

We conclude that primary single level unilateral lumbar decompressive surgery, performed without the use of a microscope, has a higher dural tear rate than the same surgery performed with the benefit of an operating microscope. Surgical time and incidence of nerve damage are also reduced by use of the microscope.

Surgeons performing hip resurfacing ante-vert and translate the femoral component anterior to maximize head/neck offset and educe impingement. The anterior femoral neck is under tensile forces during gait similarly to the superior neck [6]. This study was esigned to determine the risk of femoral neck fracture after anterior or posterior notching of the femoral neck.

Method: Fortyseven 4th generation synthetic femora were implanted with Birmingham Hip Resurfacing pros-theses (Smith & Nephew Inc. emphis, USA). Implant preparation was performed using imageless computer navigation (VectorVision SR 1.0, BrainLAB, Grmany). The virtual prosthesis was initially planned for neutral version and translated anterior, or posterior, to create the notch. The femora were fixed in a single-leg stance and tested with axial compression using a mechanical testing machine. This method enabled comparison with previously published data. The synthetic femora were prepared in 8 experimental groups:2mm and 5mm anterior notches, 2mm and 5mm posterior notches, neutral alignment with no notching (control), 5mm superior notch, 5mm anterior notch tested with the femur in 25° flexion and 5mm posterior notch tested with the femur in 25° extension We tested the femora flexed at 25° flex-ion to simulate loading as seen during stair ascent. [3] The posterior 5mm notched femoral necks were tested in extension to simulate sporting activities like running. The results were compared to the control group in neutral alignment using a one-way ANOVA:

Results: Testing Group Mean load to failure Significance Neutral (Control) 4303.09 ± 911.04N Anterior 2mm 3926.62 ± 894.17N p=0.985 Anterior 5mm 3374.64 ± 345.65N p=0.379 Posterior 2mm 4208.09 ± 1079.81N p=1.0 Posterior 5mm 3988.07 ± 728.59N p=0.995 Superior 5mm 2423.07 ± 424.16N p=0.003 Anterior 5mm in 25° flexion 3048.11 ±509.24N p=0.087 Posterior 5mm in 25° extension 3104.61±592.67N p=0.117 Both the anterior 5mm notch tested in single-leg stance and anterior notch in flexion displayed lower compressive loads to failure (3374.64N and 3048.11N). The mean load to failure value for the posterior 5mm notches in extension was 3104.62N compared to 4303.09N for the control group. Our data suggests that anterior and posterior 2mm notches are not statistically significantly weaker in axial compression. The anterior 5mm notches tend towards significance in axial compression (p=0.38) and bordered significance in flexion (p=0.087). The 5mm posterior notches were not significantly weakened in axial compression (p=0.995), but tended towards significance in extension (p=0.117). The 5mm superior notch group was significantly weaker with axial compression supporting previous data published (p=0.003). We are currently assessing offset and other variables that may reduce data spread.

Conclusion: We conclude that anterior and posterior 2mm notching of the femoral neck has no clinical implications, however 5mm anterior notches may lead to fracture. The fracture is more likely to occur with stair ascent rather than normal walking. Posterior 5mm notches are not likely to fracture with normal gait, but may fracture with higher impact activities that promote weight bearing in extension. Hip resurfacing is commonly performed on active patients and ultimately 5mm notching in the anterior or posterior cortices has clinically important implications.

Purpose: The purpose of this study was to evaluate the accuracy and precision of 3 common methods used to produce posterior tibial slope during total knee arthroplasty.

Method: The study population consisted of 110 total knee arthroplasties in 102 patients that underwent total knee arthroplasty. All procedures were performed using a standard medial parapatellar approach and all knees were replaced using the Scorpio Knee System (Stryker, Mahwah, NJ) of implants and instruments. Three treatment groups were identified retrospectively based on the method used to produce the posterior tibial slope. Group 1 used an extramedullary guide with a 0 degree cutting block tilted by placing 2 fingers between the tibia and the extramedullary guide proximally and three fingers between the tibia and guide distally to produce a 3 degree posterior slope (N=40). Group 2 used computer navigation (Stryker Navigation System, Stryker, Mahwah, NJ) to produce a 3 degree posterior slope (N=30). Group 3 used an extramedullary guide placed parallel to the anatomic axis of the tibia with a 5 degree cutting block to produce a 5 degree posterior slope (N=40). Posterior tibial slope was measured from lateral radiographs by 2 independent reviewers that were blinded to the treatment group. The reported posterior tibial slope for each sample was an average of these two measurements. Accuracy of the treatment group was evaluated using a one sample t test. Groups 1 and 2 were tested for an ideal slope of 3 degrees, and Group 3 was tested for an ideal slope of 5 degrees. An a priori sample size calculation with α=0.05 and β=0.20 showed that at least 24 samples in each treatment group were required to determine a difference of 1.5 degrees between the treatment group mean posterior tibial slope and the ideal posterior tibial slope.

Results: The mean posterior slope measurements for treatment Group 1 (4.15±3.24 degrees) and treatment Group 2 (1.60±1.62 degrees) were both significantly different than the ideal slope of 3 degrees (p=0.03 for Group 1 and p< 0.01 for Group 2). This indicates that treatment Groups 1 and 2 failed to accurately produce the ideal posterior tibial slope of 3 degrees. The mean posterior tibia slope of treatment Group 3 (5.00±2.87 degrees) was not significantly different than the ideal posterior tibial slope of 5 degrees (p=1.00). This indicates that Group 3 accurately produced the ideal tibial slope of 5 degrees.

Conclusion: The most accurate method to produce posterior tibial slope was the 5 degree cutting block with an extramedullary guide. Computer navigation had the lowest standard deviation and therefore was the most precise method. However, computer navigation was not as accurate in producing the desired posterior tibial slope as the extramedullary guide with the 5 degree cutting block. The manual method of producing tibial slope with an extramedullary guide and a 0 degree cutting block was the least precise method and not as accurate as the extramedullary guide with a 5 degree cutting block.

Posterior slope of the tibial component is an important factor in overall alignment of Total Knee Arthroplasty. The purpose of this study was to compare the accuracy and reproducibility of tibial bone cuts utilizing traditional extramedullary 0 degree and angled 5 degree cutting blocks, and computer aided navigation, in primary total knee arthroplasty.

We identified 3 groups of patients. Group one were primary total knees performed using an extramedullary 0 degree cutting block for posterior slope, group 2 were performed using an extramedullary 5 degree cutting block and the third group were performed with computer navigation. Patients in all 3 groups were age and sex matched. All operations were performed by residents or clinical fellows, under the supervision of the senior authors. Lateral digital radiographs were reviewed and posterior slope was determined in a standardized fashion. Two independent blinded researchers assessed the posterior slope using Siemens Magicweb software version VA42C_0206.

The average difference from the ideal posterior slope in navigated knees was lower than with non-navigated knees, however this was not significant (p=0.086). The average difference from the ideal posterior slope in computer navigated knees was 1.77 degrees (95% CI=1.28 to 2.26) compared to 2.37 degrees (95% CI=1.56 to 3.17) with the 5 degree cutting block and 2.70 degrees (95% CI=1.73 to 3.66) with the 0 degree block. No absolute significant difference was highlighted between the 3 groups using ANOVA testing (p=0.22).

All three techniques used to obtain ideal tibial slope were accurate. Accuracy was not increased by the use of computer navigation; however navigation resulted in less variation in outcome. The two jig based methods produced similar outcomes and either technique can be used successfully.

This study was designed to determine the risk of femoral neck fracture after anterior or posterior notching of the femoral neck. The anterior femoral neck is under tensile forces during gait similarly to the superior neck [6].

Method: Fortyseven 4th generation synthetic femora were implanted with Birmingham Hip Resurfacing pros-theses (Smith & Nephew Inc. Memphis, USA). Implant preparation was performed using imageless computer navigation (VectorVision SR 1.0, BrainLAB, Germany). The prosthesis was initially planned for neutral version and translated anterior, or posterior, to create a femoral neck notch. The femora were fixed in a single-leg stance and tested with axial compression. This method enabled comparison with previously published data. The synthetic femora were prepared in 8 experimental groups: 2mm and 5mm anterior notches, 2mm and 5mm posterior notches, neutral alignment with no notching (control), 5mm superior notch, 5mm anterior notch tested with the femur in 25° flexion and 5mm posterior notch tested with the femur in 25° extension

We tested the femora flexed at 25° flexion to simulate loading as seen during stair ascent. [3] The posterior 5mm notched femoral necks were tested in extension to simulate sporting activities like running. The results were compared to the control group in neutral alignment using a one- way ANOVA:

Results: Testing Group Mean load to failure Significance

Neutral (Control) 4303.09 ± 911.04N

The anterior 5mm notch tested in single-leg stance and anterior notch in flexion displayed lower compressive loads to failure (3374.64N and 3048.11N). The mean load to failure value for the posterior 5mm notches in extension was 3104.62N compared to 4303.09N for the control group.

Our data suggests that anterior and posterior 2mm notches are not significantly weaker in axial compression. The anterior 5mm notches was not significant in axial compression (p=0.38), but trended towards significance in flexion (p=0.087). A 5mm posterior notch was not significant. (p=0.995, p=0.117). The 5mm superior notch group was significantly weaker with axial compression supporting previous published data (p=0.003).

Conclusion: We conclude that anterior and posterior 2mm notching of the femoral neck has no clinical implications, however a 5mm anterior femoral neck notch may lead to fracture. The fracture is more likely to occur with stair ascent rather than normal walking given the reduction in strength noted after testing in flexion. Posterior 5mm notches are not likely to fracture. Hip resurfacing is commonly performed on active patients and 5mm notching of anterior cortex has clinically important implications.

Minimizing tip-apex distance has been shown to reduce clinical failure of sliding hip screws used to fix peritro-chanteric fractures. The purpose of this study was to determine if such a relationship exists for the position of the lag screw in the femoral head using a cephalomedullary device.

Methods: Thirty intact synthetic femur specimens (Model #3406, Pacific Research Laboratories, Vashon, WA) were potted into cement blocks distally for testing on an Instron 8874 (Instron, Canton, MA). A long cephalomedullary nail (Long Gamma 3 Nail, Stryker, Mahwah, NJ) was inserted into each of the femurs. An unstable four-part fracture was created, anatomically reduced, and repaired using one of 5 lag screw placements in the femoral head:

Superior (N=6),

Mechanical tests were repeated for axial, lateral and torsional stiffness. All specimens were radiographed in the anterioposterior and lateral planes and tip-apex (TAD) distance was calculated. A calcar referenced tip-apex distance (CalTAD) was also calculated.

ANOVA was used to compare means of the five treatment groups. Linear regression analysis was used to compare axial, lateral and torsional stiffness (dependant variables) to both TAD and CalTAD (independent variables).

Results: ANOVA testing proved that the mean axial (p< 0.01) and torsional stiffness (p< 0.01) between the 5 groups was significantly different, but lateral stiffness was not statistically different (p=0.494). Post hoc analysis showed that the inferior lag screw position provided significantly higher mean axial stiffness (568.14±66.9N/ mm) than superior (428.0±45.6N/mm; p< 0.01), anterior (443.2±45.4N/mm; p=0.02) and posterior (456.7±69.3N/ mm; p=0.04) lag screw positions. There was no significant difference in mean axial stiffness between inferior (568.14±66.9N/mm) and central (525.4±81.7N/mm) lag screw positions (p=0.77). Post hoc analysis revealed significantly less mean torsional stiffness for the superior lag screw position compared to other lag screw positions (p< 0.01 all 4 pairings). There were no significant correlations between TAD and axial (r=−0.33, p=0.08), lateral (r=−0.22, p=0.24) or torsional (r=0.08, p=0.69) stiffness. There were significant correlations between CalTAD and axial (r=−0.66, p< 0.01), lateral (r=−0.38, p=0.04) and torsional (r=−0.38, p=0.04) stiffness.

Discussion: Our results suggest that placement of the lag screw inferiorly in the femoral head when using a cephalomedullary nail to treat an unstable peritrochanteric fracture results in the stiffest construct in axial and torsional biomechanical testing. A simple radiographic measurement, CalTAD, provides an intraoperative method of determining optimal cephalomedullary nail lag screw position to achieve greatest construct stiffness.

We prospectively assessed a consecutive series of patients undergoing MTPJ arthroplasty with the MOJE prosthesis. All patients entered into the study were assessed preoperatively with the AOFAS 100-point Hallux Meta-tarsophalangeal-Interphalangeal Joint Scale and the range of motion was recorded. Patients were assessed on table postoperatively for range of motion (ROM) and then at 3, 12 and 24 months with AOFAS scores and ROM.

Forty-two toes (40 patients) were recruited into the study. There were 24 women and 16 men. The mean patient age on the day of surgery was 59 (range 37 to 73). 18 operations were carried out on the left hallux and 24 on the right. All operations were carried out for a diagnosis of hallux rigidus (although one patient also had hallux valgus, with an intermetatarsal angle of 24° and a hallux valgus angle of 40°).

The mean pre-op AOFAS score increased from 36.0 to 82.2 at 3 months (p< 0.001) and was 87.0 at 12 months and 84.2 at 24 months. There was no significant change in scores from 3 months onwards. Only 2 patients had a follow-up of 36 months; both of them had AOFAS scores of 95.

The mean arc of motion reduced from 70.8° on-table to 33.3° by 24 months (p< 0.001). The difference in arc of motion from 3 months to 12 months was a decrease from 45.6 to 40.0 which was borderline significant.

In 4 radiographs there was evidence of progressive loosening (figure 4). This was at 24 months in all 4 cases. For 3 of the patients the AOFAS score was 85. For the 4th patient the AOFAS score was 65

One patient had a spontaneous fusion of the toe. There were also three episodes of wound breakdown, one patient had intra-operative division of the EHL tendon that was repaired. We also noted post-operatively that: three feet developed Morton’s neuromata; one patient developed tarsometatarsal joint osteoarthritis of the great toe, one sesamoid osteoarthritis and one plantar fasciitis.

At the most recent follow-up appointment 33 out of 40 patients (82.5%) were satisfied with the results of their operation, 2 were dissatisfied (5%) and results regarding satisfaction were not available for 5 patients.

The results obtained in this paper demonstrate good, prospective, short-term results with the press-fit zirconium ceramic Moje implant. We believe that in the correct patient group good short term results can be achieved in the treatment of 1st MTPJ osteoarthritis as an alternative to fusion, particularly in those patients who are unwilling to have permanent stiffness in this joint for cosmetic or functional reasons

We present the functional and radiological results after implantation of thirty four uncemented Buechel Pappas total ankle replacements. The average follow-up period was 5 years. All consecutive uncemented Buechel Pappas total ankle replacements were performed by the senior author (MSA) between 1990 and 2005, at Corbett District General Hospital, Dudley. Thirty four ankle replacements were performed. One patient died before follow-up. All ankle replacements were performed for osteoarthritis. The patients were assessed clinically and radiologically.

Thirty four ankle replacements were performed on thirty three patients. There were eight female and twenty five male patients. One patient had died 2 years after her ankle replacement. The average age of the patients was 69(58–84). All ankle replacements were performed for osteo-arthritis. 59% had post traumatic osteoarthritis.

The average body mass index was 30 (22–37). One hundred percent of patients were happy with their ankle replacement. Fifty nine percent of patients were completely pain-free, or had occasional twinges. All patients were mobile, forty seven percent of patients required no walking aids and only six percent required 2 walking aids. One patient complained of occasional instability. One patient sustained a medial malleolar fracture intra-operatively which was fixed with two partially threaded cancellous screws. The patients were placed in a back-slab post-operatively and were non-weight bearing for a period of 6 weeks. No revision operations were performed to date.

Clinical examination revealed forty two (5–57) degrees of movement in the ankle replacement at follow-up. Radiological examination revealed forty five (0–57) degrees of movement in the ankle joint. Pearson Correlation coefficient was 0.956, p< 0.001. There was excellant correlation between clinical and radiological assessment of range of movement.

Examination of the radiographs demonstrated no evidence of gross subsidence, or lucency. One implant had a poorly positioned talar component, but so far this has been assymptomatic.

The American Foot and Ankle Society Scores were 34.6(20–56) pre-operatively compared with 76(54–100) at follow-up. This was significant improvement p< 0.001. Average scores for pain, function and alignment were 30(20–40),40(29–50) and 9(5–10).

There was one intra-operative fractured medial malleolus and two superficial wound infections (with negative swabs) treated with antibiotics.

We compare our results with others published and feel the Buechel Pappas Uncemented Total Ankle replacement performs well and is a good alternative to ankle arthrodesis preserving ankle movement and biomechanics. The results are comparable to other uncemented total ankle replacements.

To evaluate the functional outcome after anterior cruciate ligament reconstruction using bioabsorbable femoral cross-pin fixation

30 patients with chronic anterior cruciate ligament deficiency, treated by arthroscopic reconstruction using hamstring tendon autograft have been reviewed. 3.3mm bioabsorbable (polylactic acid) cross-pins (Rigidfix, Ethicon) were used for graft fixation in the femoral tunnel. Outcome assessment was performed using Lysholm score,Tegner activity level, KT-1000 Arthrometer laxity measurements, single legged hop test and SF-36 evaluation. There were 20 male and 10 female patients with an age between 17 and 45 years(mean 25.1 yrs).The mean follow up was 20.3 months (range12–29 months). The mean Tegner score was 5.5 with a range of 3 to 7. Lysholm scores averaged 88.27 with a range of 74 to 100. The range of movement in the operated knee was comparable to that in normal knee. The single legged hop was 93.4% of the unoperated side. The average KT-1000 side to side difference was 3.2 mm using a 30 lb force and 3.5mm on manual maximal load.

Biabsorbable femoral cross-pin fixation offers a viable and safe method of graft fixation in anterior cruciate ligament reconstruction and yeilds a satisfactory functional outcome.

In contrast to bony metastases, soft tissue metastases from carcinoma are rare. We reviewed all referrals to our Sarcoma Unit over an eight-year period, and found an incidence of soft tissue metastases from carcinoma of 1.4%. The most common mode of presentation was a painless soft tissue lump in a patient with an occult primary. Lung and kidney were the most frequent primary sources. Overall, prognosis was poor, with a mean survival of 9.4 months. Renal tumours however had a much better prognosis that other types of tumours. Treatment should be individualised according to the underlying disease and the prognosis. Although much rarer than primary soft tissue sarcomas, soft tissue metastases from carcinoma should remain a differential diagnosis in any patient presenting with a suspicious soft tissue lump.

Introduction: Carcinoma most commonly metastasises via the lymphatic system firstly to the regional lymph nodes and then into the general circulation. Dependent upon the primary site of tumour, metastases from carcinoma commonly occur to the lungs, liver and bone. Distant metastases to the soft tissues are rare. There are few published case series (1,2) – the majority of the literature containing only case reports. We report a series of 10 cases of soft tissue metastases from carcinoma, collected from retrospective review of the case notes of consecutive patients referred over a eight year period to our Sarcoma Unit with a soft tissue lump, suspicious of a sarcoma.

Patients and Methods: We retrospectively reviewed the case notes of consecutive patients over a eight year period (April 1995 – April 2003) referred to our Unit with a soft tissue lump, suspicious of a sarcoma. All patients underwent magnetic resonance (MR) scanning, and then trucut or open biopsy of the lesion. Dependent on the histological and MR findings, patients then underwent computer tomography (CT) of the chest and ultrasound examination of the abdomen. Included in this series were all patients with a histologically proven, soft tissue (skeletal muscle or subcutaneous tissue) metastatic carcinomatous deposit.

Demographic, diagnostic, clinical, radiological and treatment data was collected on all patients.

Results: Of the 702 referrals to our Unit over the eight-year period with a soft tissue lump suspicious of a sarcoma, 10 cases proved to be soft tissue metastases from carcinomas (incidence 1.4%). Data for the 10 patients comprising the series is shown in Table 1. Eight of the patients were male, two were female. The mean age at presentation was 68 years (range 39–85 years). Two patients presented with a painful lump, and in the other eight patients the lump was asymptomatic. The involved sites included the thigh in four cases, the arm in three cases, the back, buttock and axilla one each.In nine cases, the soft tissue lump was the presenting symptom of an occult primary carcinoma, whilst in one case (Case 4); the patient had a history of previous excision of a hypernephroma. The sources of primary carcinoma were small cell carcinoma of the lung in 4 cases, renal clear cell carcinoma in 3 cases, large bowel adenocarcinoma in 1 case, prostate 1 case, and in 1 case the primary site was unknown.

All but one patient (Case 6) underwent radiotherapy or chemotherapy or both. Case 6 presented with a soft tissue lump over the shoulder, which on biopsy was found to be metastatic adenocarcinoma of large bowel origin. CT scan of the head confirmed multiple brain metastases. He declined any treatment and died within 2 months of presentation. In total, nine of the ten patients have died of their disease. The mean duration from diagnosis of soft tissue metastasis to death was 9.4 months (range 2–31 months). The duration of survival was significantly better for metastatic carcinoma of the kidney (23 months) compared to the other carcinomas (7 months).

Discussion: The most commonly reported primary carcinomas to result in soft tissue metastases are those of the lung, kidney and colon (1–3), contrasting with those carcinomas which commonly metastasise to bone such as prostate, breast and thyroid which only very rarely metastasise to the soft tissues. Damron and Heiner (1) who reported the largest series to date however had no cases where the patients primary site of carcinoma was of renal or colon origin and suggested that these cases were over-represented in the literature. Our series differs from their findings, concurring instead with the other published literature. Histologically, the most common diagnosis is adenocarcinoma, though many have been reported (1–3).

Soft tissue metastases from carcinoma are rare, which again contrasts to bony metastases from carcinoma. Tolia and Whitmore (4) reviewed 586 patients with renal cell carcinoma, and whilst a quarter had evidence of distant metastases at presentation, no patient had soft tissue metastases. Chandler et al (5) reported on 726 patients who died following metastatic renal carcinoma; only 3 patients had soft tissue metastases, all of which were only found at autopsy. Our series, which reviewed all patients referred to our Unit over a eight year period with a soft tissue lump initially suspicious of a sarcoma, found an overall incidence of 1.4%.

Damron and Heiner (1) reported that the most common mode of presentation was a painful soft tissue lump. In our series however, the majority were painless. Whilst for most, the lump is the first sign of an occult malignancy, renal cell carcinomas tend to be different, often presenting as a solitary soft tissue deposit a few months to up to 16 years after the initial diagnosis of renal cell carcinoma has been made (1,2). Our series agreed with these findings; only in 1 case (Case 4) was there evidence of previously documented carcinoma prior to presentation with the soft tissue lump.

All patients in our series underwent pre-operative MR scans, the appearances of which were not diagnostic of metastases, though highly suggestive of malignancy. Subsequently patients underwent either Tru-cut or open biopsy which gave the definitive diagnosis. As part of the pre-operative work-up, all patients had a CT of the chest and abdominal ultrasound scanning.

Rao et al (6) reported 5 cases of soft tissue metastases from primary sarcoma, concluding that metastases in these cases were seen as late events and survival was generally poor. In our series of soft tissue metastases from carcinoma, we also found that prognosis was poor (averaging 9.4 months), especially when the primary carcinoma was lung, though the prognosis, if the primary was renal carcinoma was less bleak.

Conclusions: Although rare, soft tissue metastases from carcinoma should remain a differential diagnosis in any patient presenting with a suspicious soft tissue lump. Whilst the MR scan appearances were suggestive of malignancy, they were not diagnostic of metastases. Tru-cut or open biopsy was reliable in confirming the diagnosis of carcinoma and helpful in the detection of the possible origin of the primary. The most common primary sites were lung, kidney and bowel.