A retrospective study was conducted on 58 severe idiopathic clubfeet in 40 children treated surgically using Carroll's approach between 1984 and 1995. All had serial manipulation and casting initially. Postoperative follow-up averaged 5.5 years. Clinical and radiographic analysis using criteria modified from Porat revealed 85% good and excellent results. Carroll's approach allows for complete soft tissue releases and gives good long term results.
Key words: calcaneocuboid joint, medial incision, posterolateral incision, Porat's criteria, talocalcaneal index
Clubfoot or congenital talipes equinovarus is a deformity of multifactorial pathogenesis in which the heel is inverted, the forefoot and midfoot are supinated and adducted, and the ankle is in equinus. There is a cavus deformity with the toes at a lower level than the heel. The condition varies in severity from postural deformities to rigid deformity or joint contractures.
The management of severe resistant idiopathic clubfoot has been the topic of controversy for many years, partly due to ignorance of the aetiology and pathoanatomyl3 of this condition. Though many cases of clubfoot can be treated conservatively with serial casting and manipulation, many will not respond. The failure rate ranges from 12% to 60%,'o and is higher for the severe resistant category. The current consensus is that surgical release is the treatment of choice for severe resistant clubfoot.
There has been much debate about the correct timing for clubfoot surgery. The current trend is towards early operative treatment of severe clubfoot, as reported by Tachdjian,25 Barenfeld and Wesely,? Attenboroughl and Main et al.1 The extent of surgical management is divided into 2 main groups:
(i) radical or complete releases, aiming at full correction by one procedure. (Carroll et al.,6 Turco,26 McKay,"7 Simons24)
(ii) partial release, or limited surgery mainly confined to the posterior structures (Main et al.16).
Although the choice of surgical approach should be tailored to every individual case of clubfoot, we believe that early radical and complete release is the procedure of choice for severe idiopathic clubfoot cases. We have been using the Carroll approach for the past decade, and found it to be superior in allowing access for complete soft tissue release, especially with regard to the posterior and lateral tethers, without compromise in wound healing.
The aim of our study was to evaluate the surgical results of severe idiopathic clubfoot using the Carroll approach. The results were assessed clinically and radiographically. We also graded the postoperative results as excellent, good and poor using criteria modified from Porat et al.23
MATERIALS AND METHODS
A retrospective study was conducted on 40 children (58 feet) with idiopathic severe clubfeet treated at the National University Hospital during the period between 1984 and 1995. Clubfoot secondary to neuromuscular causes (e.g. cerebral palsy, myelomeningocele and arthrogryposis) were excluded from the study. The severity of the clubfoot was graded as mild, moderate, severe or very severe (Dimeglio et al.9). The cases included in our paper were all severe and very severe cases, i.e. resistant feet that were either partially reducible or almost irreducible. Figures l(a) and (b) illustrate a typical example of bilateral severe idiopathic clubfeet.
There were 28 boys and 12 girls in our study, with the male to female ratio being 2.3:1. The ethnic composition was as follows: Chinese (22 cases: 55%); Malay (9 cases: 22.5%); Indian (7 cases: 17.5%); Others (2 cases: 5%). The deformity was bilateral in 18 cases (45%), with the male to female ratio being 2:1. There were 22 cases of unilateral clubfoot (55%), with the right foot being affected in 11 cases and left foot in 11 cases. The duration of follow-up ranged from one year to 12 years, with an average follow-up of 5.5 years.
All patients were screened for other abnormalities. There were 4 cases who presented with additional anomalies. One patient had Fallot's Tetralogy where a right Blalock-Taussig shunt was performed at 3 months of age. The second patient had cleft palate (corrective surgery done at 9 months of age), constriction bands of the right index and ring fingers (corrective surgery done) and syndactyly affecting the right middle 3 toes (corrective surgery done). The third patient had constriction bands affecting the right second to fourth digits (corrective operation done). The fourth case had nail patella syndrome.
Treatment of these patients usually began as soon as they were referred to our orthopaedic department. All except 3 cases were referred soon after birth. Initial management of all these cases was conservative. The deformities were managed by weekly manipulation and below knee casting. Plasters were changed every week and no attempt was made to dorsiflex the foot until the forefoot adduction and heel varus had been corrected. Early soft tissue release was recommended if there was no correction of the deformity after 3 to 4 months of plaster treatment. Usually, the surgery was performed between one to 2 months after the decision for surgery was made. The average age at operation was 6.7 months in our series. This included children with clubfeet who presented late.
All patients had preoperative radiographs (AP and lateral clubfoot views) performed one day prior to surgery. Radiographic assessment consisted of AP talocalcaneal angle, lateral talocalcaneal angle and Kite's angle (combined AP and lateral talocalcaneal angles).
The patients were intubated and positioned prone on a standard operating table. Tourniquets were used throughout the operation to ensure a bloodless operating field. All cases of clubfeet were surgically reconstructed using the Carroll approach, via a medial incision and a posterior lateral incision.57
The child was discharged on the second postoperative day and returned for casting one week later. The limb was treated in a below-knee cast for a period of 12 weeks. The cast was changed every 4 weeks. The Kirschner wire was removed at the end of 8 weeks. Dennis-Browne splints set at 450 of external rotation were then worn till the child started to walk. The bar was also bent at 300 to evert the feet. In our series, the average duration of DennisBrowne splints prescribed was 6.5 months. Postoperative evaluation
Clinical (questionnaire and clinical examination)
In the first postoperative year, the child was followed up every month until the cast was discontinued, usually at 3 months after surgery, and thereafter every 2 to 3 months. In the second and third postoperative years, the child was reviewed on a 6-monthly basis. The child was reviewed yearly after the third postoperative year. The clinical review included the shape of the foot, any residual deformity (in particular the heel - varus/valgus or equinus; forefoot adduction or supination), mobility of the subtalar joints, range of motion of both ankle and subtalar joints, gait (including presence or absence of heel and toe walking), leg length discrepancy, foot and calf sizes. The child and parents were questioned about the presence of pain (no pain; pain with walking; pain with sports), functional capability (no limit; limitation in walking; limitation in functional activities) and type of shoewear (normal vs special shoes). Clinical photographs were also taken at this time. Figures 2 to 5 are postoperative photographs of a girl with bilateral severe clubfeet taken 10 years after surgery. There were no obvious deformities present, and the patient had no functional limitations or pain.
Radiographic studies were also performed preoperatively and postoperatively. Standard AP and lateral clubfoot views were taken. The AP and lateral talocalcaneal angles were measured and added up to give Kite's angle (Talocalcaneal index).
There were no cases of immediate post-operative complications. All wounds healed primarily and there were no cases of wound infection or breakdown.
Residual foot deformities (Table 1)
Varus heel was found in 12 feet (20.7%). All of these feet had less than 50 varus. Valgus heel (>50) was found in 2 feet, implying an over correction of the hindfoot. None of these cases required any orthosis or special shoes, nor exhibited signs of functional disability. Neutral to 5" valgus heel was found in the remainder, i.e. 44 feet (75.9%).
Forefoot adduction (>50) was found in 16 feet (27.6%). Of these 16 feet, 2 had reoperations done. Both had extensive soft tissue release performed, and one had decancellation of cuboid done. 42 feet (72.4%) had their forefoot in a neutral to 5 adduction alignment.
Forefoot supination was present in 15 feet (25.9%). Of these 15 feet, one had reoperation done. 43 feet (74.1%) had no supination of the forefoot.
There were 3 cases of equinus heel (5.2%). One of them had reoperation done. 55 feet (94.8%) were plantigrade.
Postoperative range of motion (Table 2)
The subtalar joints were mobile in all 58 feet. The average ankle dorsiflexion was 14.30, with a range of -10o (fixed equinus) to 200 (full range of motion). The average ankle plantar flexion was 28.90, with a range of 10o to 40o. The average inversion at the subtalar joint was 25.30, with a range of 13o to 30o. The average eversion at the subtalar joint was 16.70, with a range of 5o to 20o. The average ankle dorsiflexion and plantarflexion for the contralateral uninvolved foot (unilateral cases) were 200 and 400 respectively. The mean inversion and eversion at the subtalar joints for the contralateral uninvolved foot (unilateral cases) were 300 and 200 respectively.
Clinical gait observation
There were 3 cases of abnormal gait, all of them due to fixed equinus (i.e. non plantigrade) deformity. 37 patients (92.5%) exhibited normal gait pattern. With regards to the ability to perform heel walking and toe walking, 3 patients were unable to heel walk due to fixed equinus deformity. All the patients were able to walk on their toes.
Leg length discrepancy, foot/calf sizes (Table 3)
There were 7 cases (17.5%) of leg length discrepancy. Five of these had a leg length discrepancy of 0.5 cm, whereas the other 2 had discrepancies of 1 cm and 2 cm each. With regards to differing foot and calf sizes (measured as foot length along the bimalleolar axis and midcalf circumference respectively), we noted 21 patients had unequal foot sizes (11 cases had 0.5 cm difference in foot length; 6 cases had 1 cm difference in foot length; one case had 1.5 cm difference in foot length; 3 cases had 2 cm difference in foot length). 23 patients presented with unequal calf sizes (13 cases had 0.5 cm difference in midcalf circumference; 6 cases had 1 cm difference in midcalf circumference; 2 cases had 1.5 cm difference in midcalf circumference; 2 cases had 2 cm difference in midcalf circumference). The foot length and calf circumference did not differ by more than 3 cm in all the cases. Since all cases were severe or very severe, it was not surprising to note that 21 out of 22 unilateral clubfoot cases had unequal foot sizes and all 22 unilateral clubfoot cases had differing midcalf sizes.
Questionnaire (Table 4)
36 patients never experienced pain. One patient had occasional foot pain on walking. Three patients experienced pain with sporting activities. Functional ability was full in 36 patients with 4 cases experiencing some limitation in daily functional activities like sports. None had limitation in walking. With regards to footwear, 2 patients had to wear special outflare shoes. The others had no problems wearing normal shoes.
Radiographic analysis (Table 5)
The average preoperative AP talocalcaneal angle was 14.20 (range 0-40O). The average postoperative AP talocalcaneal angle was 30.2o (range 10o-70o). The mean preoperative lateral talocalcaneal angle was 11.6 (range Oo-24o). The mean postoperative lateral talocalcaneal angle was 27.10 (range 13o-45o). The average combined preoperative talocalcaneal index or Kite's angle was 25.80 (range 8-50o). The average postoperative Kite's angle was 57.3 (range 38o-107o). The postoperative talocalcaneal angles still lagged behind those values obtained from normal uninvolved feet (p
Figures 6 to 9 illustrate a typical example of a boy with right sided unilateral clubfoot. Preoperative radiographs (Figs. 6 and 7) revealed almost parallel orientation of the talus and calcaneus on the lateral view. After extensive soft tissue release via Carroll's approach, the AP talocalcaneal index improved to 31o (Fig. 8) and the lateral talocalcaneal index to 450 (Fig. 9), giving a Kite angle of 760.
Classification of surgical results
We grouped our surgical results into excellent, good and poor, based on both clinical and radiographic criteria modified from l'orat et al.13 (Tables 6a,b). The criteria were foot and ankle mobility, gait pattern, foot shape, lower limb growth and radiography. Six patients (15%) had poor postoperative results. The first patient had fixed equinus deformity, hindfoot varus, forefoot adduction (>5) / supination and abnormal gait. The second patient had fixed equinus deformity, forefoot adduction and abnormal gait. The third case had forefoot adduction/supination and talocalcaneal index of 36o (below 40o). Two cases had hindfoot varus and forefoot adduction/supination. The last patient had fixed equinus deformity, hindfoot varus, forefoot adduction and abnormal gait. Good and excellent results were achieved in 34 patients (85%).
DISCUSSION Congenital clubfoot is a complex deformity. The goal of treatment is to obtain a painless, normal looking foot with good mobility and function. Treatment of clubfoot has evolved from prolonged manipulation and casting, minimal surgery such as posterior releases,'4 to radical complete releases.,24 Most orthopaedic surgeons would agree that the initial treatment of a clubfoot patient should be conservative. The method of choice is manipulation and plaster cast application at weekly intervals.5ls,zo Any clubfoot that cannot be corrected or only partially corrected with 3 to 4 months of serial manipulation and casting is considered resistant and surgery is usually contemplated.
Modern treatment principles of resistant clubfoot were conceptualized by Bradford3 in 1892. Brockman4 described his extensive soft tissue release (in 2 stages) in 1930. The importance of early surgical release was emphasized by Attenborough.1 Turco's approach26 was a major turning point in the surgical management of this difficult condition. Carroll6 refined the technique by emphasizing the release of the lateral tether, including the calcaneocuboid joint, through 2 incisions. The Cincinnati incision was later described by Crawford et al.8 to allow for complete posteromedial and lateral release.
The comparison of surgical results of clubfoot is difficult due to 3 reasons: (1) varying severity of clubfoot, (2) different treatment methods and (3) different postoperative assessment protocols. In our department, all cases were initially treated with weekly manipulation and casting. The timing of surgery for resistant and severe cases is controversial, ranging from newborn'4,19 to 2 or 3 years of age. The current consensus would advocate operation by 3 to 6 months.' In our series, the average age at surgery was 6.7 months. Most resistant cases were operated on at 3 to 4 months of age. Those who were operated on later either presented late or had a longer period of serial manipulation and casting.
During the past decade, we have been using the Carroll approach, which offers good accessibility for soft tissue release and anatomical reconstruction. An important finding in the pathoanatomy in clubfoot surgery is the state of the posterolateral tissues.2 The calcaneofibular ligament is a thick cord that forms the main element of the lateral tether. The posterior talofibular ligament and the posterolateral capsule, together with the calcaneofibular ligament keep the heel in equinus position. Carroll's approach allows good access to these structures, unlike the Turco incision where these structures are not as easily approached.
The Carroll approach emphasizes the need to release the calcaneocuboid joint.6 This is important for proper alignment of the forefoot on the hindfoot. In Turco's approach the calcaneocuboid joint is not dealt with at all, which might contribute to the recurrence of forefoot adduction. We had 16 surgically treated clubfeet which had persistent significant (>50) adduction. Many of these (11) were due to residual adduction deformity at the tarsometatarsal joints.
Although the Cincinnati incisions is widely used, we believe it does not allow as complete a posterior release as Carroll's approach would allow, especially in severe and late cases that present with severe equinus deformity. The circumferential incision in the Cincinnati approach makes it difficult to perform an adequate tendo Archilles lengthening, an important step in the correction of a severe equinus deformity. The posterolateral incision of the Carroll approach enables us to do tendo Archilles lengthening easily, even in older children with very severe equinus deformities.
The importance of complete subtalar release was advocated by McKayl7 and Simons.24 In our experience, we found that this release is necessary to enable us to derotate the talus and calcaneus at the subtalar joint. Complete subtalar reconstruction depends on the completeness of the peritalar release. The Carroll approach provides good access to the subtalar joint from both the medial and postero-lateral incisions through which the release of the talonavicular and calcaneocuboid joints is performed.
In our experience, we did not encounter problems with wound healing and scar contracture using the Carroll approach. In contrast, many venous channels are divided in the long curved incision of Turco2" which might affect skin healing. In addition, the hypertrophic and contracted scar might lead to later forefoot adduction.
There are many ways to assess postoperative results, as exemplified by the many scoring systems available in the literature. Broadly they can be subjective and/or objective outcome assessments. None of these assessments is perfect. As mentioned earlier, comparison of the results of our study with the other studies is difficult because of differing evaluation schemes and the varying duration of follow-up. Nevertheless, it is important to understand that people with clubfoot would never be functionally better off than those *vithout the deformity. Periodic long term follow-up is essential due to the possibility of deterioration of results with time.'2 Growth increases the severity of existing deformities.
Based on the literature, early complete soft tissue releases demonstrate varying success rates. Franke and Hein" reported 94% success rate with early soft tissue release. Turco26 reported excellent or good results for 125 (84%) of 149 feet with a follow-up of 2 to 15 years. Assessment was based on clinical (including range of motion) and radiographic criteria, but not on function or symptoms. In 1985, Simons24 reported satisfactory results for 72% of clubfeet treated with complete subtalar release compared with 50% of clubfeet treated with less extensive release.
Although the corrected clubfeet in our series looked and functioned fairly normally, they were different from the normal contralateral feet in most objective evaluation parameters. The clubfeet had a reduced range of motion as measured with a hand held goniometer. We did not find markedly restricted plantarflexion in our series, as reported by Porat and Kaplan.22 Care was taken to avoid overzealous lengthening of the tendo Archilles intraoperatively. Both the lengthened Archilles and tibialis posterior tendons were sutured with the foot in plantigrade position. This would allow correction of the equinus deformity yet avoid the problem of weakening plantarflexion power.
Residual foot deformities were present in a number of patients, yet only a few exhibited functional disability. The most common deformity was forefoot adduction (16 out of 58 feet). Many of these may be due to inadequate release of the calcaneocuboid joints, but the majority were due to residual adduction deformity at the tarso-metatarsal joints. With follow-up, the adduction deformity improved with time. There were 15 feet with residual supination. One probable explanation was that the tibialis anterior muscle was over-acting against the relatively weaker peronei. Two patients required subsequent tibialis anterior transfer to address this issue. The cause of varus/valgus heel may be due to error in intraoperative judgment, leading to undercorrection or over-correction of the hindfoot. In our opinion, the most severe residual deformity would be a persistent equinus heel. We considered the presence of equinus deformity a poor surgical result. Fortunately, only 3 out of 58 feet had equinus deformity.
Differing foot and calf sizes were present in clubfeet, especially the unilateral ones where the difference was much more obvious. None differed by more than 3cm. It is important to emphasize to the parents that smaller foot and calf sizes in these children would persist.
Our clubfeet had a decreased Kite's angle compared to the normal contralateral feet. On the other hand, the values of bimalleolar foot angle between the 2 were quite close. Many of the clubfeet were visually plantigrade despite radiographic abnormalities, similar to the study by Laaveg and Ponseti.ls In that study, they claimed 74% good and excellent results functionally. However, only 11.5% of these cases showed that the lateral talocalcaneal index had been corrected.
Subjective results based on pain, function and gait observation must not be ignored in our quest for objective results in terms of numbers and angles. In fact, they might provide a better measure of the final long term outcome. In our study, very few had residual pain (3 cases) or functional limitation (4 cases).
In our study, clinical and radiographic criteria were also used to grade the surgical results. Good and excellent results were achieved in 85% of cases, which compares favourably with other reported series.,26 We used Porat's23 criteria, with some modifications, to grade our clinical results, because we found this system easy to use. Other grading systems were more cumbersome and complicated, especially those with scores attached to each parameter. Undue importance might be attached to certain parameters, which in turn could bias the overall surgical result. We would like to emphasize the importance of long term follow-up to detect any clinical deterioration and persistent deformities that may require further surgery. The final outcome of clubfoot surgery is best measured at the end of growth.
Carroll's dual incision approach offers good exposure to the medial and posterolateral soft tissue structures around the foot and ankle, allowing adequate release of all these tethers. In our study, 85% of surgically treated severe clubfeet had good to excellent results. Long term follow-up is essential to assess persistent deformities which may need additional surgery.
The authors would like to thank Ms Grace Lee for her assistance in gait analysis and pedobarography, Mr Tan Boon Kiat for the photography and Ms Sarojeni and Ms Liah for typing the manuscript.
1. Attenborough CG. Early posterior soft-tissue release in severe congenital talipes equinovarus. Clin Orthop 1972, 84:71-8. 2. Barenfeld PA, Wesely MS. Surgical treatment of congenital clubfoot. Clin Orthop 1972, 84:79-87. 3. Bradford EH. Operative treatment of resistant clubfoot. Trans Am Orthop Assoc 1892, A-5:183. 4. Brockman EP. Modern methods of treatment of clubfoot. Br Med J 1937, 2:572. 5. Carroll NC. Congenital clubfoot: Pathoanatomy and treatment. In: Instr Course Lect, American Academy of Orthopaedic Surgeons 1987, 36:117-21.
Carroll NC, McMurtry R, Leete SF. The pathoanatomy of congenital clubfoot. Orthop Clin North Am 1978, 9:225-32. Carroll NC. Pathoanatomy and surgical treatment of resistant clubfoot. In: Instr Course Lect, American Academy of Orthopaedic Surgeons 1988, 37:43-106.
Crawford AH, Marxen JL, Osterfeld DL. The Cincinnati incision: A comprehensive approach for surgical procedures of the foot and ankle in childhood. J Bone Joint Surg 1982, 64-A:1355-8.
Dimeglio A, Bensahel H, Souchet PH, Mazeau PH, Bonnet F. Classification of clubfoot. Part B.J Pediatr Orthop 1995,4:12936.
De Rossa GP, Stepro D. Results of posteromedial release for the resistant clubfoot. J Pediatr Orthop 1986, 6:590-5. Franke J, Hein G. Our experience with the early operative treatment of congenital clubfoot. J Pediatr Orthop 1988, 8:26-30. Green ADL, Lloyd-Roberts GC. The results of early posterior release in resistant clul)feet. J Bone Joint Surg 1985, 67-B:58893.
13. Herzenberg IE, Carroll NC, Christofersen MR, Lee EH, White S, Munroe R. Clubfoot analysis with three-dimensional computer modeling. J Pediatr Orthop 1988, 8:257-62.
14. Hudson I, Catterall A. Posterolateral release for resistant clubfoot. J Bone Joint Surg 1994, 76-B:281-4. 15. Laaveg SI, Ponseti IV. Long term results of treatment of congenital clubfoot. J Bone Joint Surg 1980, 62-A:23-31. 16. Main BJ, Crider Rj, Pol KM. The results of early operation in tal ipes equi novarus: A preliminary report. J BoneJoint Surg 1977, 59-B:337-41.
17. Mct Dl. New conceptof and approach to clubfoot treatment. Section Correction tl. Conect+on of the clubfoot.IPediatri1983, 3:1 -21.
18. Nather A, Bose K. Conservative and surgical treatment of clubfoot. J Pediatr Orthop 1987, 7:42-8. 19. Pons JG, Dimeglio A. Neonatal surgery in clubfoot. Orthop Clin North Am 1978, 9:233--40. 20. Ponseti IV, Smoley EN. Congenital clubfoot: The results of treatment. J Bone Joint Surg 1963, 45-A:261-75. 21. Porat S. Evaluation of surgical treatment in resistant clubfoot: A comparison of the Turco, Carroll and Cincinnati approaches.
In: The Clubfoot. New York: Springer-Verlag, 1994, 479-88.
22. Porat S, Kaplan L. Critical analysis of results in clubfeet treated surgically along the Norris Carroll approach: Seven years of experience. J Pediatr Orthop 1989, 9:137-43.
23. Porat S, Milgrom C, Bentley G. The history of treatment of congenital clubfoot at the Royal Liverpool Children's Hospital: Improvement of results by early intensive posteromedial release. J Pediatr Orthop 1984, 4:331-8. 24. Simons GW. Complete subtalar release in clubfeet. Parts I and II. J Bone Joint Surg 1985, 67-A:1044-65. 25. Tachdjian MO. Pediatric Orthopaedics. Vol. 1. Philadelphia: WB Saunders, 1972,1274-322. 26. Turco V). Resistant congenital clubfoot: One stage posteromedial release with internal fixation: A follow-up report of a fifteen year experience. J Bone Joint Surg 1979, 61 -A:805-14.
Hwan-Tak Hee and Eng-Hin Lee
Department of Orthopaedic Surgery, National University Hospital, Singapore.
Address correspondence and reprint requests to: Dr Eng-Hin Lee, Department of Orthopaedic Surgery, National University Hospital, 5 Lower Kent Ridge Road, Singapore 119074.
Copyright Western Pacific Orthopaedic Association Dec 1997
Provided by ProQuest Information and Learning Company. All rights Reserved