Abstract
Aims
To systematically review the efficacy of split tendon transfer surgery on gait-related outcomes for children and adolescents with cerebral palsy (CP) and spastic equinovarus foot deformity.
Methods
Five databases (CENTRAL, CINAHL, PubMed, Embase, Web of Science) were systematically screened for studies investigating split tibialis anterior or split tibialis posterior tendon transfer for spastic equinovarus foot deformity, with gait-related outcomes (published pre-September 2022). Study quality and evidence were assessed using the Methodological Index for Non-Randomized Studies, the Risk of Bias In Non-Randomized Studies of Interventions, and the Grading of Recommendations Assessment, Development and Evaluation.
Results
Overall, 17 studies (566 feet) were included: 13 studies used clinical grading criteria to report a postoperative ‘success’ of 87% (75% to 100%), 14 reported on orthotic use with 88% reduced postoperative use, and one study reported on ankle kinematics improvements. Ten studies reported post-surgical complications at a rate of 11/390 feet (2.8%), but 84 feet (14.8%) had recurrent varus (68 feet, 12%) or occurrence of valgus (16 feet, 2.8%). Only one study included a patient-reported outcome measure (pain).
Conclusion
Split tendon transfers are an effective treatment for children and youth with CP and spastic equinovarus foot deformities. Clinical data presented can be used for future study designs; a more standardized functional and patient-focused approach to evaluating outcomes of surgical intervention of gait may be warranted.
Cite this article: Bone Jt Open 2023;4(5):283–298.
Take home message
Split tendon transfers are effective at restoring a functional plantigrade foot, improving gait function, and reducing the need for orthotics in about 87% of children and youth with cerebral palsy.
12% of patients have recurrence of varus deformity and 3% have development of the opposite deformity at mean follow-up of five years.
About one in ten patients undergo further foot surgery, providing useful data in counselling parents about expected outcomes.
Future standardized patient-reported and performance-related evaluation measures are warranted to assess impact on pain, function, and quality of life.
Introduction
Spastic equinovarus (EQV) foot deformities are a common foot deformity in CP.1 These foot deformities and biomechanical changes are not present at birth, but evolve in individuals with spastic CP as they grow.1 Overactivity of either the tibialis anterior (TA) or the tibialis posterior (TP) muscles, or both, can overpower the peroneal muscles and will contribute to the forefoot and midfoot being pulled into varus and supination (overactive TA), and the hindfoot into varus (overactive TP). This leads to increased weightbearing on the lateral aspect of the foot, with reduced foot contact surface area and base of support, thus creating instability during gait.2 The underlying TA and TP muscle overactivity is often accompanied by a shortening of the calf muscles, which pulls the ankle joint into an equinus/plantar flexed position.3 Combined equinus and varus further reduce foot contact during gait phases, hindering foot clearance during swing phase and increasing the likelihood of falls, and difficulty with orthotic fitting and shoe wear.4-6
Split anterior tibialis tendon transfer (SPLATT) and split posterior tibialis tendon transfer (SPOTT) are common tendon transfer surgical procedures used to correct muscle imbalance in spastic EQV in CP. These procedures work by altering the muscle-tendon-bone attachments to balance actions of the foot inverter and evertor muscles and, in doing so, acting to restore foot stability. When the varus is mostly present within the forefoot during swing phase (with high TA activity), a SPLATT to the lateral side of the foot (either the peroneus tertius (PT),7,8 peroneus brevis (PB),7-9 or the cuboid bone)10-14 is indicated, aiming to neutralize the opposing varus force while enhancing dorsiflexion.4,9 By contrast, when the impairment is more within the stance phase of gait with the hindfoot in varus and equinus (and an overactive TP), a SPOTT to the peroneus brevis (or the lateral cuneiform) on the lateral side of the foot is indicated.15 This is considered preferable to TP lengthening, to minimize altering the capacity of the muscle to produce power through range, an undesired product of lengthening.4,5,13 To a large degree, the actions of TA and TP muscle activity will overlap due to the phenomenon of segmental linkage, and for this reason both gait analysis and electromyographic assessments provide valuable adjuncts for operative planning.16-18 Preoperative clinical assessments also aid in determining if concomitant soft-tissue procedures are needed, such as lengthening of the Achilles tendon, or Gastrocnemius or TP.19
Not all patients with EQV should be managed with the same approach, with seminal work by Chang et al19 highlighting that the type of CP and preoperative ambulatory status play an important factor in the success of split tendon transfers. Historically, studies reported findings with grouped cohorts featuring multiple surgical interventions, thus compromising the conclusions of isolated split tendon transfer surgery.20-23 As such, the impact on gait and functional outcomes is difficult to distinguish. The age at operation has also varied between studies; early conservative management of foot deformities means that the first operation for equinovarus can be in adolescence, or when the individual is skeletally mature. The aim of this systematic review was to evaluate the literature to determine the moderate- to long-term efficacy of isolated SPLATT/SPOTT surgery for children and youth with CP and spastic EQV, with a focus on gait-related outcomes. The United Nations definition of youth was applied, with an age of 15 to 24 years.24
Methods
The protocol (PROSPERO registration CRD42021276809) followed the steps and guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).25
Search strategy
Five electronic databases were comprehensively searched (by JHO) on 27 August 2022: the Cochrane Central Register of Controlled Trials (CENTRAL, the Cochrane library), CINAHL, Embase - OVID, PubMed, and Web of Science with search strategies comprising Medical Subject Headings (MESH)/keywords relating to CP, equinovarus, and split anterior/posterior tibialis tendon transfer (Supplementary 1). References of included studies were hand-searched for additional articles meeting the criteria.
Study eligibility
Two authors (JHO, SAW) used Covidence software (Australia) to independently review titles and abstracts, with records meeting the inclusion criteria progressing to full-text screening. Discrepancies between authors were resolved though discussion and through consultation with a third author (SS). Records unavailable in English were translated using Google Translate.
Inclusion criteria were: 1) participants with a diagnosis of CP and spastic equinovarus foot deformity, age range from three to 24 years at time of surgery, irrespective of functional ability level; 2) randomized control trials (RCT), quasi-RCT, case series, retrospective or prospective cohort studies, case-control and case studies with > ten participants; 3) intervention included split tibial tendon transfers; 4) reported follow-up period > 24 months (mean) and/or used instrumented gait analysis with a mean follow-up > 12 months; and 5) at least one outcome measure of interest. These outcome measures included: clinical grading criteria such as Green et al,4 Kling et al,5 and Hoffer et al10 (Table I), variables of gait from 3D gait analysis, gait speed, distance, endurance, gross motor function or mobility, clinical examinations/classifications including ankle and subtalar range of motion (ROM), radiological measurements and angles, reported complications, rate of equinovarus recurrence and revision surgery, need for orthotics, and patient-reported outcomes (e.g. pain).
Table I.
An outline of descriptions of the clinical criteria determined by Kling et al,5 Green et al,4 and Hoffer et al.10
| Grade | Kling et al’s5 criteria (established 1985) | Green et al’s4 criteria (established 1983) | Hoffer et al’s10 criteria (established 1974) |
|---|---|---|---|
| Excellent |
|
|
|
| Good |
|
|
|
| Poor | Feet:
|
Feet:
|
Feet:
|
-
Excellent and good results are rated successful according to Kling et al5 and Green et al4 whereas those whose feet had recurrent equinovarus or varus deformity, overcorrection, under-correction, or equinus greater than 5°, and orthotics were required during gait, were considered poor/failed results and subsequently required further surgery.
Studies were excluded if < ten participants with CP and EQV (n = 3),26-28 or if the participants had idiopathic congenital talipes equinovarus; concomitant surgical procedures (including foot osteotomies and/or any knee/hip surgery), or previous tibial tendon transfers or lower limb bone/neurosurgical interventions. We made efforts to contact corresponding authors to obtain data from studies where data for subgroups of participants meeting the review eligibility criteria were not presented. Studies were subsequently excluded if we were unable to obtain a response/data from the corresponding authors, however they were included if these data were less than 5% of their overall cohort.4,7,29
Data extraction and reporting
Data were extracted using a purpose-developed extraction form, including available information on study and participant characteristics, variables related to the intervention and rehabilitation, and the reported outcome measures, which were categorized by their corresponding domain (JHO, SAW) within the International Classification of Functioning Disability and Health (ICF).30 Data were tabulated to facilitate synthesis and interpretation of the results. Calculated averages for clinical graded criteria, need for orthotics, complications, recurrences, and revisions were recorded to provide an overall synthesis of results. Due to confounding variables of age, Gross Motor Function Classification Scale (GMFCS), and topography in addition to heterogeneity between reported outcomes, we were unable to complete subgroup analysis and meta-regression.
Data from studies only presenting preoperative information are not included within the synthesis of data, nor are outcomes of walking distance, endurance, anatomical, and foot switch measures due to heterogeneous methods, measurements, and infrequent/lack of reporting. The intention of the review to also evaluate outcomes of gross motor function/mobility measures was not possible, as these measures were not included within the studies included for review.
Quality assessment
Two reviewers (JHO, SAW) independently assessed quality using the Methodological Index for Non-Randomized Studies (MINORS),31 risk of bias using Risk of bias In Non-Randomized Studies of Interventions (ROBINS-I),32 and quality of evidence using GRADE33 for clinical criteria outcomes. Any disagreements were resolved through discussion. A third reviewer (NSS) resolved any further discrepancies (full scoring provided in the Supplementary Material).
Results
Search results and study quality
The original search yielded 546 records, of which 309 duplicates were removed, 245 underwent screening of title and abstract, 57 going through full-text review, and 20 subsequently included for review (Figure 1). Seven publications required additional data/information to confirm eligibility and for reporting,7,9,14,19,34-36 from which four were received,7,9,14,36 and three were subsequently excluded (data unavailable from authors),19,34,35 resulting in a total of 17 studies included for data extraction. Of these 17 studies, authors of nine studies reported or supplied subgroup data for the subgroups of participants with CP meeting our inclusion criteria.7,9,13-15,17,29,36,37 Two studies had participants undergoing our target surgical intervention as part of multilevel soft-tissue surgery.7,14
Fig. 1
PRISMA search strategy for the present systematic review. F/U, follow-up.
Scores on the MINORS ranged from 6 to 12 (Supplementary Table ii). The majority of studies (11/17) were graded as moderate-risk according to the ROBINS-I assessment, mostly owing to retrospective analyses, non-blinded assessors, and use of subjective clinical measures (Supplementary Table iii). The GRADE summary of evidence (Supplementary Tables iv to vi) illustrates varied certainty of evidence between very low (measures of ambulation), to low (clinical grading criteria, orthotics use), to moderate certainty (maximum dorsiflex during swing, radiological measures, foot kinematics, spatiotemporal parameters, pain (Likert scale), difficulty in shoe wear (Likert scale), early complications, and recurrences). The GRADE quality of evidence for split tibial tendon transfers restoring a functional plantigrade foot for children and youth with CP and EQV deformity (Supplementary Table iv) was moderately confident in the estimated effect: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. By contrast, low quality of evidence was determined in relation to gait functional outcomes (Supplementary Table v), largely due to the serious concerns of risk of bias due to subjective, unblinded measures and lack of quantitative objective measures in the majority of studies. Equally, there were serious concerns overall in the imprecision due to the small effect sizes in the studies which did highlight quantitative measures. As such, our confidence in this effect estimate is limited: the true effect may be substantially different. The overall quality of the evidence for split tendon transfers reducing pain postoperatively was moderate (Supplementary Table vi). This was reflected within the GRADE domains, with no serious concerns of risk of bias, inconsistency, indirectness, or publication bias. However, due to the low estimate of effect,9 imprecision was rated as a serious concern.
Characteristics of included studies
The 17 studies reported on 506 patients (200 females: 195 males, sex not reported in five studies)4,10,11,17,37 and 566 feet (Table II). All 17 studies were non-randomized, seven were retrospective follow-up case series, and ten were indicated to be case series. Within the 17 studies, five drew comparison of outcomes between type of foot deformity (i.e. equinovarus vs varus),17,29 intervention (i.e. modified SPLATT vs standard SPLATT or SPLATT vs SPOTT),13,15 and type of CP (i.e. hemiplegia vs diplegia).37
Table II.
Study and participant characteristics of all 17 studies included in the systematic review, organized chronologically.
| Study | Study design | Total patients, n; F:M | Mean age, yrs (range)* | CP type | Total feet, n† | Deformity pre-index reported, n‡ |
|---|---|---|---|---|---|---|
| Green et al, 19834 | CS | 16; NR | 6.2 (4.5 to 10) | H13, Q3 | 16 | 14 EQV flexible, 2 fixed hindfoot varus deformity |
| Kling et al, 19855 | CS | 31; 19:12 | 8.0 (3.5 to 18) | H19, D5, Q7 | 37 | 37 EQV/flexible |
| Hoffer et al, 198510 | CS | 21; NR | NR (5 to 26) | NR | 27 | 6 EQV flexible, 21 EQV with ‘some’ fixed deformity |
| Barnes and Herring, 199111 | CS | 12; NR | 7.8 (12 to 15) | H9, D1, M2 | 12 | 9 varus/flexible, 3 fixed |
| Synder et al, 199338 | CS | 21; 13:8 | 12.9 (4.8 to 21) | H8, D4, Q4, M1 | 21 | 20 EQV, 1 fixed hindfoot varus deformity |
| Saji et al, 199329 | CS | 18; 10:8 | 8.2 (4 to 14.5) | H10, D7, Q1 | 23 | 10 hindfoot and forefoot varus without equinus, 13 hindfoot and forefoot varus with equinus, 14 fixed, and the remaining (n = 9) flexible |
| Mulier et al, 199537 | CS | 12; NR | 5.5 (4 to 9)§ | H10, D2 | 14 | 14 EQV/flexible |
| O'Byrne et al, 199739 | CS | 16; 10:6 | 10.1 (5.6 to 12.7) | H7, D9 | 16 | 16 EQV/flexible |
| Scott and Scarborough, 200617 | CS | 25; NR | 10.8 (5 to 17) | H25 | 25 | 8 EQV, 17 hindfoot varus, all flexible |
| Vlachou and Dimitriadis, 201013 | RCS | 48; 33:12¶ | 12.4 (9 to 18) | H32, D12, Q4 | 52 | 34 EQV mid- and forefoot, 18 Equinus hindfoot varus/ all flexible |
| Ahmed et al, 201112 | CS | 20; 12:8 | 8.0 (6 to 10) | H15, D4, Q1 | 25 | 25 EQV/flexible |
| Limpaphayom et al, 201514 | RCS | 22; 6:16 | 7.35 (4.24 to 13.02) | H6, D16 | 32 | 32 EQV/ flexible |
| Aleksić et al, 202015 | RCS | 124; 38:86 | 11.0 (IQR 6.00 to 14.25) | H65, D38, Q15, P22, T6 | 146 | 146 EQV/flexible |
| Lullo et al, 20207 | RCS | 37; 18:19 | 8.8 (3.3) | NR | 37 | 37 varus/3 feet plantarflexion contractures (passive dorsiflexion < 0°), 34 forefeet and hindfeet could be everted passively to at least 0° |
| Sarikaya et al, 20208 | RCS | 26; 14:12 | 8.1 (5 to 14) | NR | 26 | 26 EQV/flexible |
| Wong et al, 20219 | RCS | 52; 25:27 | 9.7 (6.1 to 14.11) | H52 | 52 | 52 EQV/combined flexible and fixed |
| Dussa et al, 202136 | RCS | 5; 2:3 | 14.4 (7.0 to 21.0) | H4, D1 | 5 | 5 EQV flexible varus foot deformity |
-
*
Mean age at index surgery.
-
†
Feet in the study which meet the review's inclusion criteria.
-
‡
Fixed is defined as not passively correctable to neutral, while flexible is defined as passively correctable to neutral.
-
§
Mulier et al do not account for the average age for the specific subgroup cohort which fit the inclusion criteria.
-
¶
Vlachou et al do not account for sex in 3 patients who underwent joint SPLATT and SPOTT.
-
CP, cerebral palsy; CS, case series; D, diplegia; EQV, equinovarus; H, hemiplegia; IQR, interquartile range; NR, not reported; P, paraplegia; Q, quadriplegia; RCS, retrospective case series; T, tetraplegia.
Around half of the studies (n = 9/17) described participants as having ‘flexible’ EQV foot deformities, and seven reported some participants with fixed deformities (four studies of which were conducted prior to the year of 1993).4,7,9-11,29,36,38 Two studies out of the 17 focused on hindfoot and forefoot varus deformities,7,11 and three compared groups between EQV and varus foot deformities.13,17,29 Out of the total 506 patients, 275 were classified as hemiplegic, 99 diplegic, 35 quadriplegic, two monoplegic, 22 paraplegic, and five tetraplegic. Three studies did not report CP topography (i.e. n = 68 patient topography unaccounted for).7,8,10 All studies included in the review were aligned with treating foot muscle imbalance in spastic CP muscle tone, however the tone abnormality was not graded in severity prior to surgical intervention.
Indications for surgery were an EQV/varus gait pattern, and cohorts were often described as ambulant, with or without aids, or had the “potential to become ambulant” postoperatively. Only five studies reported participants’ preoperative GMFCS level (GMFCS I, n = 49, II, n = 117, III, n = 22, IV, n = 26, V, n = 21, and n = 5 combined level I/II).7-9,14,36 These specific levels were not reported in studies pre-dating 2011, however brief descriptions of ambulation were provided.
Surgical interventions
Participant ages at time of the index intervention ranged from 3.3 to 26 years, with a calculated mean age of nine years, however the majority (14 of 17 studies, 82%) were between 3.3 and 17 years. Data from one study included a patient aged 26 years who could not be removed from the group analysis.10 Although this participant was older than this review’s eligibility criteria, we elected to retain the study within the review. Eight studies (205 feet) implemented the SPLATT intervention,7-14 of which three studies were published in 2020 or, more recently,7-9 implementing the SPLATT to Peroneus Brevis/Peroneus Tertius approach versus the SPLATT to cuboid approach (used in studies prior to 2015). Ten studies implemented SPOTT (353 feet),4,5,12,13,15,17,29,37-39 with the majority reporting the SPOTT to Peroneus Brevis surgical approach. Out of these studies, two reported either SPLATT or SPOTT intervention groups depended on the clinical presentation,12,13 and one had a subgroup combining both procedures during index operation (n = 3 feet).13 One study combined both interventions for all participants (n = 5 feet).36 The majority of the studies reported participants as having no previous operations, however Kling et al5 and Mulier et al37 included data from 14 participants (out of a total of 43 participants) with soft-tissue procedures prior to index operation which were unable to be differentiated in the data. A small number (< 5%) of participants within the study by Lullo et al7 had soft-tissue or bony procedures.
Three studies reported use of above-knee casts for at least six weeks with periods of restricted weightbearing,5,12,36 with nine reporting the use of below-knee casts and earlier weightbearing.7,8,10,11,14,15,17,29,39 Four studies4,5,13,37 described a combination of above- and below-knee casts. Follow-up times varied widely from one to 16 years, with seven studies5,9,10,13,14,29,40 reporting outcomes at a mean of five years or more, and six studies over seven years post-surgery.5,9,10,13,29,40
Complications, recurrences, and revisions
Ten studies reported on post-surgical complications (presence or absence of),5,7-11,14,15,36,38 with four reporting the presence of complications. An overall reported early complications rate of 2.8% (n = 11/390 feet) was apparent: one necrosis of the skin flap, one superficial wound infection, two pressure sores over the heel,5 four cast sores,7,9 two pain/spasm,9 and one described as “wound detachment.”8
Recurrence of EQV foot deformity postoperatively was reported in all studies. The total overall (varus and equinus) foot deformity recurrences recorded were 68 out of 566 feet (12%), with ten studies reporting < three recurrences.4,5,8,10,11,17,29,37-39 Post-surgical development of the opposite deformity – plano or equinovalgus was reported by four studies and in 16 feet (2.8%).7,9,29,36
Revision surgery was reported for 67 out of the 84 equinovarus or valgus postoperative occurrences. Overall, 11.8% of feet required a second surgical procedure (n = 67/566).
Outcome measures
All studies reported gait-related outcomes categorized in the ICF body structure and function domain,30 and three reported measures at the level of activity (Table III).8,14,40 The most commonly reported gait-related outcome measures included the use of gait analysis, gait observation, ambulation, range of motion (ROM), and clinical grading tools. Table IV outlines the results of overall clinical function at postoperative follow-up. Table V and Table VI include outcomes from 16 studies which presented pre- and postoperative data including assessments at both timepoints. One study did not include pre- and postoperative data.29
Table III.
Details on surgical intervention, post-surgical conditions (i.e. rehabilitation, the use of casting or bracing etc), and the follow-up assessment. Included outcome measures are categorized by ICF domains Body structure and Function (B), Activity (A), and Participation (P).
| Study | Soft-tissue release at index | Surgical approach | Casting, protocols/rehabilitation | Mean follow-up (range) |
Complications/revisions/recurrence | Outcome tools/measures | ICF domain/s |
|---|---|---|---|---|---|---|---|
| Green et al, 19834 | 15 feet sliding heel-cord lengthening, 1 foot calcaneus osteotomy* |
SPOTT to PB | 4/52 long cast with knee extension and foot neutral then BK for walking for 4/52. If the patient can actively DF to neutral then no brace or splint required. | Minimum 2 yrs |
Complications: NR Recurrences/revisions: 1 (6.3%) closing wedge osteotomy a year later |
Clinical examination,† gait observation via video analysis, EMG, pressure switches, Green criteria | B |
| Kling et al, 19855 | Heel cord lengthening” was reported in “most cases”. Where this was not done at index it was required a year later. Preoperatively, 6 feet had heel cord lengthening 2/6 also had TP lengthening. |
SPOTT to PB | 6/52 long cast with knee slight flexion and foot in neural position - BK for 2/52. PT as soon as comfortable - WB encouraged | 8 yrs (4 to 14) |
Complications: Total 4 (13%) (1 (2.7%) necrosis, 1 (2.7%) wound infection, 2 (5.4%) pressure sores) Recurrences: 3 (8.15%) varus Revision: Total 3 (8.15%) (1 (2.7%) heel cord lengthening (8 yrs old), 1 (2.7%) triple arthrodesis (10 yrs old), 1 (2.7%) lengthening of TP (12 yrs old)) |
Clinical examination, gait observation, Kling criteria, radiographs, photos. WB AP radiographs of foot, ankle, and calcaneus |
B |
| Hoffer et al, 198510 | 13 feet had TAL, 8 feet had TP lengthening |
SPLATT to the cuboid | 6/52 BK for WB then AFO for 6/52 | Minimum 10 yrs |
Complications: None (0%) Recurrences: 1 (3.7%) varus Revision: 1 (3.7%) calcaneal osteotomy |
Clinical examination, gait observation, PROM, EMG, foot contact studies/foot switch, functional ambulatory ability, orthotic need | B |
| Barnes et al, 199111 | 6 feet had lengthening, 2 feet had plantar fasciotomy |
SPLATT - cuboid with intramuscular lengthening TP | 6/52 BK brace with WB - no AFO required after in management | 5.5 yrs (2.25 to 8.75) |
Complications: None (0%) Recurrences: 3 (25%) varus (aged 8, 14, and 11 yrs). Revisions: Total 2 (16.6%) 1 (8.3%) triple arthrodesis, 1 (8.3%) TP transfer |
Clinical examination, gait observation via video analysis, Kling criteria | B |
| Synder et al, 199338 | All had TAL | SPOTT to PB | 6/52 long cast - AFO 6/12 | 3.4 yrs (2.5 to 6.3) |
Complications: None (0%) Recurrences: 3 (14.2%) varus Revision: 1 (4.8) requiring a Dwyer osteotomy |
Clinical examination, gait observation, WB roentgenograms, Kling criteria | B |
| Saji et al, 199329 | 13 feet had release of TA, 1 foot had an osteotomy* |
SPOTT to lateral cuneiform via interosseous membrane | 6/52 BK with heel in slight valgus and foot plantigrade. WB at 3/52 postop - short leg caliper with posterior strap for 6/12 to 1 yr postop | 8.4 yrs (4 to 14.5) |
Complications: NR Recurrence: 1 (4.3%) in the varus group due to valgus deformity Revision: 1 (4.3%) tenotomy 4 yrs later |
Clinical examination, gait observation via video analysis/photography, AP/lateral radiography in standing, Kling criteria, EMG | B |
| Mulier et al, 199537 | All had subcutaneous sliding heel cord lengthening (preop 8 patients had heel cord lengthening) | SPOTT to PB via interosseous membrane | 3/52 long cast with knee in extension and foot neutral position then 3/52 BK for walking | 37 mths§ (35 to 37) |
Complications: NR Recurrence: 2 (14.2%) varus hemiplegia group (5 yrs/11 yrs) Revision: 2 (14.2%) Due to technical error and they went on to have 1 (7.1%) triple arthrodesis and 1 (7.1%) Dwyer osteotomy |
Clinical examination, Kling criteria and Green criteria combined, gait observation, radiographs, and photos |
B |
| O'Byrne et al, 199739 | 13 feet had tendocalcaneus lengthening | SPOTT to Peroneus brevis | 6/52 BK in neutral foot position | 1 yr (NR) |
Complications: NR Recurrences/revisions: None (0%) |
Clinical examination, gait observation, Green criteria, 3DGA (kinematic analysis of ROM ankle stance phase, and max DF during swing phase) | B |
| Scott et al, 200617 | 18 feet had TAL lengthening | SPOTT-NR | 6/52 BK/WB - AFO 6/12 | 4.8 yrs (NR) |
Complications: NR Recurrences/revisions: n = 3 (12%) varus (mean age 12.9 yrs) |
Clinical examination, 3DGA (kinematic analysis of equinus in stance and drop foot in swing), EMG, Kling criteria | B |
| Vlachou et al, 201013 | Group I: 11 feet had plantar soft-tissue release, 8 had transcutaneous flexor tenotomies, and 5 had a Jones procedure. Group II: 23 feet transcutaneous flexor tenotomies, f= 18 achilles cord lengthenings, f = 15 plantar soft-tissue releases, f = 5 Jones, 2 feet extension tendons transfer to the metatarsals |
Gr I: SPLATT - cuboid gr I (11 feet forefoot and midfoot inversion) Gr II: SPOTT - PB (38 feet hindfoot varus) Three patients had combined SPLATT and SPOTT |
4/52 long cast with knee extended/foot neutral at 4/52 WB as tolerated - 2/52 BK. If patient able to DF to neutral then no brace is required | 7.8 yrs (4 to 14) |
Complications: NR Recurrences 4 (7.7%) varus recurrences Revision: 4 (7.7%) calcaneum fusion 16 to 18 months post-index |
Clinical examination, Kling criteria, Hoffer criteria, AP/lateral WB radiographs | B |
| Ahmed et al, 201112 | 25 feet had percutaneous lengthening tendocalcaneous, 14 feet had plantar fascia release | SPLATT to cuboid (forefoot varus and EQV);‡ SPOTT-IM-Lateral cuneiform (hindfoot varus and EQV) | 6/52 long cast POP ankle neutral - 3/52, touch down WB - AFO 6/12 | 3.5 yrs (2 to 4) |
Complications: Not reported Recurrences: 4 (16%) with marked TA shortening, 2 (8.0%) varus, 2 (8.0%) equinus Revision: 4 (16%) calcaneum fusion 18 mths post-index operation |
Clinical examination, ROM, Kling criteria | B |
| Limpaphayom et al, 201514 | All had lengthening of TA, PT, FDL, FHL | SPLATT - cuboid | 12/52 BK/WB | 5.5 yrs§ (1.1 to 16) |
Complications: None (0%) Recurrences: 4 (12.5%) varus Revision: 4 (12.5%) |
Clinical examination, Kling criteria,n GMFCS | B, A |
| Akeksic et al, 202040 | 73 feet had Hokes procedure, 25 feet had strayerst, 1 foot had Achilles elongation, 1 foot had Achilles elongation, 6 feet had Z-plasty, 1 had tenotomy of FD muscle, 5 feet had the Hoke TAL and elongation of FD muscle, 2 feet had elongation of FD muscle |
SPOTT 41 feet: to PB (modified group) and remaining part of TP tendon lengthened, standard (green techniques) SPOTT to PB 105 feet | 6/52 BK/WB cast, rehab protocol mentioned but no details | 8.0 yrs (IQR 6 to 11) |
Complications: None (0%) Recurrence: 29 (19.8%), 26 (17.8%) in standard intervention group, 3 (2.0%) modified group. Revision: 19 (13.0%). Triple arthrodesis in 16 (15.2%) feet in the standard group, 3 (7.3%) in the modified group. |
Clinical examination, gait observation, Kling criteria, GMFCS | B, A |
| Lullo et al, 20207 | 16 feet had Gastrocnemius recession, 10 feet had TAL, 18 feet had PTTL, 15 feet had SPOTT Pre-index 2 feet had osteotomies* |
SPLATT- PB/tertius | 6/52 BK/WB as tolerated then part time bracing (at least 6 to 8 hrs a day) 6/12 postoperatively | 4.4 yrs (4.0) |
Complications: 1 (2.7%) cast sore Recurrence: 11 (29.7%) (8 (21.6%) varus (average age 5.1 yrs and 3.2 years post-intervention)) n = 3 (8.1%) valgus, (average age 11.4 yrs and 3.7 yrs post intervention) Revisions: 9 (24.3%) (7 varus (18.9%), 2 (5.4%) valgus) |
Medical records, clinical examination, 3DGA (Vicon-kinematic analysis DF during swing phase), PROM, Modified clavien-Dindo complication rating scale, postoperative recurrences |
B |
| Sarikayai et al, 20208 | 10 feet had Gastroc tenotomy, 3 feet had gastrocnemius tenotomy and plantar fasciotomy, 2 feet had achilloplasty, 2 feet had achilloplasty and plantar fasciotomy, 1 foot had a plantar fasciotomy, 1foot had Flexor carpi ulnaris tenotomy |
SPLATT to peroneus tertius and 2 feet SPLATT to cuboid | 6/52 BK in neutral position | 28.8 mths§ (24 to 42) |
Complications: 1 (4%) wound detachment Recurrence/revisions: none (0%) |
Kling criteria, GMFCS, clinical examination, gait observation, medical records |
B, A |
| Wong et al, 20219 | All had IML-TP and SPLATT-PB. 14 feet had B toxin to gastrocnemius, 5 feet had strayers, 2 feet had strayers with solealfascial lengthening, 31 feet had a modified vulpius gastrocsoleus recession, and 11 had white slide lengthening of the AT |
SPLATT-PB | Hinged AFO for 12 mths. Hinge closed in 44 patients to protect calf lengthening surgery for 3 to 6 mths | 7 yrs (3.3 to 10.2) |
Complications: 5 (9.6%) (3 (5.8%) cast sores, 2 (3.8%) pain/spasm) Recurrence: 12 (23.1%) (3 (5.7%) recurrent varus, 2 due to persistent varus (10.3 yrs, 14.8 yrs, 16 yrs), 9 (17.3%) valgus) Revisions: 12¶ (23.1%) (3 (5.7%) SMO, 2 (3.8%) Evans, 3 (5.7%) vulpius, 2 (3.8%) SPLATT tenotomy, 2 (3.8%) CCFx) |
Clinical examination, gait observation via video analysis, AP/lateral radiographs in WB, Likert pain scale, Likert shoe wear, medical records | B |
| Dussa et al, 202136 | 3 feet had the Baumann procedure | SPLATT-cuboid, SPOTT-PB, all feet were combined, SPLATT/SPOTT | 21.4 mths (13 to 37) |
Complications: None (0%) Recurrences: 3 (60%) valgus Revision: 1 (20%) |
Clinical examination, AP/lateral radiological analysis, kinematic gait analysis | B |
-
Complications = early postoperative complications.
-
No previous operations reported if not stated above.
-
*
< 5 % bony surgery previous/at index.
-
†
Clinical examination involved: observed gait pattern - foot alignment and balance and muscle control, active range of motion, passive range of motion, observation of any callus formation, or wear and tear on shoes.
-
‡
Does not subgroup number of SPLATT and SPOTT in study.
-
§
Specific subgroup data not available.
-
¶
Study reported a delay in the older patients accessing surgery relayed to delayed referral to the service with prolonged cycles of injections of Botulinum Toxin A.
-
AFO, ankle foot orthosis; AP, anterior-posterior; AT, Achilles tendon; BK, below-knee cast (short leg cast); CCFx, calcaneocuboid shortening fusion plus medial deep plantar release; DF, dorsiflexion; 3DGA, 3D gait analysis; EMG, electromyography; EQV, equinovarus; f, number of feet; FDL, flexor digitorum longus; FHL, flexor hallucis longus; GMFCS, Gross Motor Function Classification Scale; ICF, International Classification of Functioning Disability and Health ; IML-TP, intramuscular tibialis posterior lengthening; IQR, interquartile range; NR, not reported; NWB, non-weightbearing; PB, peroneus brevis; POP, plaster of paris; PT, posterior tibialis; PTTL, posterior tibial tendon lengthening; PWB, partial weightbearing; ROM, range of motion; SMO, supramalleolar osteotomy of tibia; SPLATT, split anterior tibialis tendon transfer; SPOTT, split posterior tibialis tendon transfer; TA, tendo-Achilles; TAL, tendo-Achilles lengthening; TP, tibialis posterior; WB, weightbearing.
Table IV.
Clinical scoring outcomes (number of feet meeting criteria and % of overall feet included in the study) and subgroup comparison analysis post-split tendon transfer of the 13 studies using such criteria.
| Study | Total feet | Excellent, n (%) | Good, n (%) | Successful, n (%)† | Poor, n (%)‡ |
|---|---|---|---|---|---|
| Kling criteria | |||||
| Kling et al, 19855 | 37 | 30 (81) | 4 (11) | 34 (92) | 3 (8) |
| Barnes et al, 199111 | 12 | 9 (75) | 0 (0) | 9 (75) | 3 (25) |
| Synder et al, 199338 | 18 | 12 (67) | 3 (17) | 15 (83) | 3 (17) |
| Saji et al, 1993 29 | |||||
| EQV group | 13 | 12 (92) | 1 (8) | 13 (100) | 0 (0) |
| Varus group | 10 | 2 (20) | 7 (70) | 9 (90) | 1 (10) |
| Mulier et al, 1995 37 | |||||
| Hemi gr | 12 | 4 (33) | 6 (50) | 10 (83) | 2 (16) |
| Diplegic gr | 2 | 0 (0) | 2 (100) | 2 (100) | 0 (0) |
| Scott et al, 2006 17 | |||||
| EQV group | 8 | 8 (100) | 0 (0) | 8 (100) | 0 (0) |
| Varus group | 17 | 9 (53) | 5 (29) | 14 (82) | 3 (18) |
| Vlachou et al, 2010 13 | |||||
| SPLATT Gr I* | 11 | 8 (73) | 3 (27) | 11 (100) | 0 (0) |
| SPOTT Gr II | 38 | 20 (53) | 14 (37) | 34 (90) | 4 (10) |
| SPLATT and SPOTT* | 3 | 2 (67) | 1 (33) | 3 (100) | 0 (0) |
| Ahmed et al, 201112 | 25 | 14 (56) | 7 (28) | 21 (84) | 4 (16) |
| Limpaphayom et al, 201514 | 32 | 23 (72) | 5 (15.5) | 28 (87.5) | 4 (12.5) |
| Sarikayai et al, 20208 | 26 | 21 (81) | 5 (19) | 26 (100) | 0 (0) |
| Akeksic et al, 2020 40 | |||||
| Standard gr | 105 | 45 (43) | 34 (32) | 79 (75) | 26 (25) |
| Modified gr | 41 | 25 (70) | 13 (32) | 38 (93) | 3 (7) |
| Green criteria | |||||
| Green et al, 19834 | 16 | 11 (69) | 4 (25) | 15 (94) | 1 (6) |
| O'Byrne et al, 199739 | 16 | 13 (81) | 3 (19) | 16 (100) | 0 (0) |
| Total | 442 | 298 (66) | 116 (26) | 385 (87) | 57 (13) |
-
*
Hoffer criteria converted to Kling.
-
†
Successful outcomes equal total of excellent and good, able to walk with a plantar grade foot < 5° varus, valgus, or equinus, normal shoe wear, no prescribed orthotics during gait, and patients and carers are satisfied.
-
‡
Poor outcomes were classified as failed outcomes with recurrent varus/equinus or development of valgus/calcaneovalgus.
-
EQV, equinovarus.
Table V.
Collated pre- and postoperative data.
| Study | Orthotic use | Ambulation/Gross motor status | ||||
|---|---|---|---|---|---|---|
| Preop, n | Postop, n | Improved, n (%) | Preop | Postop | Change | |
| Green et al, 19834 | 16 | 0 | 16 (100) | |||
| Kling et al, 19855 | 36 | 4 | 32 (89) | 28 mobile unaided | 1 required crutches | |
| Hoffer et al, 198510 | 27* | 1 | 26 (96) | 19 community walkers, 2 non-ambulators | Improvements in the 2 non-ambulatory | |
| Barnes et al, 199111 | 2 | 0 | 2 (100) | 12/12 community walkers, 1 with crutches | No change | |
| Synder et al, 199338 | 21 | 3 | 18 (86) | 18 ambulant, 1 able to stand, 2 wheelchair-bound | 18 ambulant, 1 crutches household-ambulant, 2 non-ambulant improved DF and able to wear regular shoes | |
| Mulier et al, 199537 | 13 | 2 | 11 (85) | |||
| Scott et al, 200617 | 25* | 0 | 25 (100) | |||
| Vlachou et al, 201013 | 48* | 4 | 44 (92) | |||
| Ahmed et al, 201112 | 25* | 4 | 21 (84) | |||
| Limpaphayom et al, 201514 | 32* | 4 | 28 (88) | GMFCS I = 3, II = 16, III = 3 | GMFCS I = 18, II = 3, III = 1 | 15 improved to I, 2 improved to II |
| Sarikayai et al, 20208 | 26 | 0 | 26 (100) | GMFCS II = 17, III = 3, IV = 6 | GMFCS, II = 21, III = 4, IV = 1 | 4 improved to II, 4 improved to III/1 to II |
| Akeksic et al, 202040 | 146* | 29 | 117 (80) | |||
| Wong et al, 20219 | 52 | 6 | 46 (88) | Pain = 2.8 (SD 1.2)† Shoe wear = 4.1 (SD 1.2)† |
Pain = 1.6 (SD 0.9) Shoe wear = 1.6 (SD 0.9) |
Data indicate reduced pain and improved shoe wear‡ |
| O'Byrne et al, 199739 | 16 | 0 | 16 (100) | |||
| Total change in orthotic use | 428 (88) | |||||
-
*
Assumed all required orthotics preoperatively, however not clearly reported.
-
†
Likert scale (0 to 9) where 0 indicates no pain or no difficulties with shoe-wear or brace-wear, and 9 indicates either severe pain or being unable to use a brace or regular shoes.
-
‡
No formal statistics available for subgroup data.
-
DF, dorsiflexion; GMFCS, Gross Motor Function Classifications Scale; PF, plantar flexion; SD, standard deviation.
Table VI.
Collated pre- and postoperative data: radiological outcomes/gait analysis. Values are presented as means and ranges.
| Study | Preop | Postop | Change |
|---|---|---|---|
| O'Byrne et al, 199739 | Max DF = -9.1° (PF) | Max DF = 8.6 (DF) | Improved DF* |
| (11.5° DF to 34.13° PF) | (3.7° DF to 16.9°DF) | Positioned closer to ‘normal’ mean TFPA of 15°* | |
| TFPA = 8° | TFPA = 18° | ||
| Lullo et al, 20207† | Max DF in swing = 2.5° (DF) | Max DF in swing = -3.5° (PF) | p = 0.18 |
| Mean DF in swing = -3.2° (PF) | Mean DF in swing = -8.14° (PF) | p = 0.14 | |
| Time of max DF (% swing) = 54.6 | Time of max DF (% swing) = 80.5 | p = 0.22 | |
| PROM forefoot eversion = 17.7° (SD 9.4°) | PROM forefoot eversion = 18.2° (10.4°) | p = 0.94 | |
| PROM hindfoot eversion = 9.2° (SD 6.6°) | PROM hindfoot eversion = 8.7° (SD 6.8°) | p = 0.85 | |
| Wong et al, 20219 | Radiological comparison | Radiological comparison | |
| TCA = 88° | TCA = 65.7° | 22.3°, 95% CI 16.1° to 27.3°, p < 0.001 | |
| LTCA = 20.3° | LTCA = 41° | 20.7°, 95% CI -23.5° to -19.5°, p < 0.001 | |
| NCO = 2.2% | NCO = 42.9% | 40.7%, 95% CI -46.7% to -34.5%, p < 0.001 | |
| TNCA = -26.8° | TNCA = 23.2° | 50°, 95% CI -56.2° to -46.7°, p < 0.001 | |
| LT first MTA = -9.8° | LT first MTA = 13.6° | 23.4°, 95% CI -27° to -20.2°, p < 0.001 | |
| APT first MTA = -20.2° | APT first MTA = 7.9° | 28.1°, 95% CI -36.7° to -26.9°, p < 0.001 | |
| MSA = 30.1° | MSA = 11.1° | 19°, 95% CI 16.9° to 23.2°, p < 0.001 | |
| Dussa et al, 202136 | Radiological comparison, ° | Radiological comparison, ° | |
| LTCA = 43.1 (33.5 to 54.1) | LTCA = 43.9 (28.7 to 52.8) | p = 0.51 | |
| LT 1st MTA = 3.9 (-12.3 to 23.5) | LT 1st MTA = 8.4 (-11.2 to 24.6) | p = 0.10 | |
| LCI = 15.2 (0.0 to 24.8) | LCI = 14.4 (2.9 to 23.7) | p = 0.47 | |
| LTNA = 7.8 (-4.3 to 21.3) | LTNA = 6.8 (-7.5 to 18.9) | p = 0.59 | |
| LNCA = −1.3 (-13.8 to 14.1) | LNCA = 5.1 (-10.9 to 33.9) | p = 0.13 | |
| APTCA = 10.0 (-4.5 to 26.8) | APTCA = 13.7 (4.5 to 26.3) | p = 0.08 | |
| APTN = -15.5 (-38.1 to 17.7) | APTN = -1.2 (-24.2 to 24.0) | p < 0.05 | |
| APT 1st MTA = -26.4 (-50.7 to -36.5) | APT 1st MTA = -11.2 (-36.5 to 13.2) | p < 0.05 | |
| APT 2nd MTA = -20.7 (-44.6 to 17.6) | APT 2nd MTA = -2.9 (-29.3 to 28.3) | p < 0.05 | |
| APC 4th MTA = -21.2 (-36.8 to 1.6) | APC 4th MTA = -10.1 (-24.4 to 14.0) | p < 0.05 | |
| Clinical assessment | Clinical assessment | ||
| Passive ankle DF = 0.4 °(-10 to -5) | Passive ankle DF = 1.9 ° (-5 to 5) | p = 0.17 | |
| Passive ankle PF = 36.9 ° (15 to 50) | Passive ankle PF = 22.7 ° (10 to 45) | p < 0.05 | |
| ‡ Ankle DF strength = 30 (0 to 5) | Ankle DF strength = 2.7 (0 to 5) | p = 0.34 | |
| Ankle PF strength = 2.8 (1 to 5) | Ankle PF strength = 2.5 (0 to 4) | p = 0.27 | |
| Spatiotemporal parameters | Spatiotemporal parameters | ||
| Velocity = 34.0% (20.0 to 47.0) | Velocity = 32.8% (16.0 to 45.0) | p = 0.40 | |
| Step length = 61.5 (34 to 79) | Step length = 60.1 (33 to 71) | p = 0.40 | |
| Cadence = 54.5% (48 to 62) | Cadence = 54.1% (41 to 61) | p = 0.79 | |
| Foot kinematics: hindfoot to tibia, ° | Foot kinematics: hindfoot to tibia, ° | ||
| Mean eversion stance = 8.1 (0.3 to 15.7) | Mean eversion stance = -4.7 (-10.9 to 8.9) | p < 0.05 | |
| Mean eversion swing = 7.5 (-0.3 to 14.9) | Mean eversion swing = -4.7 (-12.3 to 9.3) | p < 0.05 | |
| Flexion ROM stance = 14.2 (7.9 to 22.3)§ | Flexion ROM stance = 15.9 (7.9 to 22.1) | p = 0.03 | |
| Eversion ROM stance = 8.0 (5.3 to 12.2)¶ | Eversion ROM stance = 6.8 (3.4 to 14.5) | p = 0.14 | |
| Rotation ROM stance = 20.6 (10.2 to 40.5) | Rotation ROM stance = 20.5 (12.4 to 38.5) | p = 0.94 | |
| Foot kinematics: midfoot to forefoot, ° | Foot kinematics: midfoot to forefoot, ° | ||
| Midfoot supination stance = -3.8 (-14.6 to 5.6) | Midfoot supination stance = 5.6 (-4.9 to 17.1) | p < 0.05 | |
| Midfoot adduction stance = 17.5 (7.8 to 28.8) | Midfoot adduction stance = 8.8 (-2.4 to 25.8) | p < 0.05 | |
| Flexion ROM stance = 9.4 (4.9 to 18.4) | Flexion ROM stance = 12.3 (4.5 to 24.5) | p = 0.03 | |
| Supination ROM stance = 9.9 (3.0 to 13.9) | Supination ROM stance = 9.1 (6.06 to 15.1) | p = 0.38 | |
| Adduction ROM stance = 6.1 (3.2 to 10.0) | Adduction ROM stance = 5.5 (2.6 to 8.0) | p = 0.40 |
-
*
No formal statistics available for subgroup data.
-
†
Swing phase gait kinematic parameters (patients without concomitant gastrocnemius recession or tendo-Achilles lengthening only n = 18).
-
‡
Strength measured on a scale of 0 to 5.
-
§
Hindfoot to tibia flexion ROM stance.
-
¶
Hindfoot to tibia eversion ROM stance.
-
APC 4th MTA, anterior-posterior calcaneal 4th metatarsal angle; APTCA, anterior-posterior talocalcaneal angle; APTN, anterior-posterior talonavicular; APT 2nd MTA, anterior-posterior talo-2nd metatarsal angle; APT 1st MTA, anteroposterior talo-1st metatarsal angle; CI, confidence interval; DF, dorsiflexion; GMFCS, Gross Motor Function Classifications Scale; LCI, lateral calcaneal inclination; LNCA, lateral navicular cuneiform angle; LTCA, lateral talocalcaneal angle; LTNA, lateral talo navicular angle; LT 1st MTA, lateral talo-1st metatarsal angle; Max DF, maximal dorsiflexion during swing phase measured in plantar flexed position ranging from dorsiflexion to plantar flexion; MSA, metatarsal stacking angle; NCO, navicular cuboid overlap; PF, plantar flexion; PROM, passive range of motion; ROM, range of motion; SD, standard deviation; TCA, tibiocalcaneal angle; TFPA, transverse foot-placement angle; TNCA, talonavicular coverage angle.
Overall clinical function at postoperative follow-up
Studies commonly used a clinical grading tool at final follow-up such as Kling (n = 11), Green (n = 2), or Hoffers (n = 1), rating ‘successes’ by a combination of ‘excellent’ and ‘good’. The mean ‘success’ rate was 87%, ranging between 75% to 100% (mean follow-up time 6.3 years) of which six studies had a success rate greater than 90% (Table V).4,5,8,13,29,39 Aleksić et al15 reported higher rates of surgical success (defined by Kling) in patients with higher levels of gross motor function prior to the surgery: GMFCS I = 100% success, II = 94.8%, III = 69.8%, IV = 9.1%, however the participant numbers in each classification level were not reported. Aleksić et al15 also reported success rates for patients with hemiplegia (58/65 (89%)), diplegia (29/38 (76%)), and quadriplegia (2/13 (15%)).
Changes in orthotic needs and changes in shoe wear
A total of 14 studies directly reported on orthotics use, with postoperative use during gait reduced by 88% (428/485 feet). All studies reported orthotic usage improvements over 80% following surgery (i.e. no orthotics prescribed postoperatively). One study reported on patient-reported levels of difficulty in shoe-wearing which reduced postoperatively.9
Changes in foot alignment and passive range of motion
Preoperative and postoperative values for weightbearing radiological measures were reported in two studies.9,36 Compared with standardized normative measures reported in the literature,41 the preoperative radiological values of the included cohorts differ substantially. Postoperative values changed significantly in all reported radiological measures from Wong et al,9 and in 4/5 anteroposterior radiological parameters reported by Dussa et al.36 Passive assessments of ROM showed no significant changes in forefoot or hindfoot eversion,7 or dorsiflexion,36 postoperatively.
Changes in ankle motion in gait
Two studies7,39 used instrumented gait analysis pre- and postoperatively. O’Byrne et al39 used a single-segment foot model, with data indicating a (non-significant) postoperative change from a more plantar-flexed position to a dorsiflexed position through swing. Using a multisegmented foot model, data from Dussa et al,36 showed statistically significant improvements in ankle eversion through gait (reductions through stance and swing), increases in hindfoot-to-tibia flexion ROM, and in supination and adduction of the midfoot-to-forefoot segment.
Pain
Only one study measured (patient reported) pain,9 with low preoperative levels that reduced further postoperatively, however we were unable to complete a statistical analysis of the change in pain data from the provided data.
Discussion
This review evaluated the outcome of split tibial tendon transfers for spastic EQV feet in children and youth with CP. Overall, these transfers had an 87% success rate in improving outcomes related to gait, based on clinical grading tools, postoperative orthotic requirements, radiological changes, and gait studies. However, there was a 12% recurrence of EQV foot deformity and a 2.8% occurrence of the opposite valgus deformity, with > 80% of the subsequent recurrent foot deformities undergoing revision surgery. Overall, the review found that the quality of reporting of the evidence was generally poor, with low- to moderate-certainty of evidence to support a successful outcome, and it was not possible to identify factors that predicted better outcomes from surgery. Gaps in evidence were also apparent, with a lack of functional objective measures relating to activity and participation domains of the ICF, and an absence of patient-/caregiver-reported outcomes.
The most common approach to evaluating outcomes of SPLATT and SPOTT interventions was through the use of one of three clinical grading tools (the Kling,5 the Green,4 or the Hoffer10) - determining a collectively high ‘success’ rate for patients. These clinical grading tools capture a range of key clinical elements relevant to the goals of the intervention, relating to gait (i.e. plantigrade foot alignment, reduced orthotics need, shoe-wear) and foot deformity, with the Kling criteria also factoring in patient satisfaction. These findings were further supported by Wong et al,9 who also reported that the surgery was effective in relieving symptoms of pain and difficulties with shoe-wear and brace-wear. Although easy to use, the Kling, Green, and Hoffer clinical grading tools have not been tested for reliability and ability to detect change. Different scoring systems such as the Patient-Reported Outcome Measurement Information System (PROMIS) item banks and the core outcome set for lower limb surgery in children with CP may prove to be better choices for future studies.42,43 Given the lack of patient-reported measures, future clinical studies would also benefit from including a validated measure of pain as well as considering use of the Gait Outcome Assessment List (GOAL) (evaluating gait priorities and functional mobility for ambulant children with CP)44 and the Oxford Ankle and Foot Questionnaire for Children (OxAFQ-C) (assessing subjective wellbeing for child patients).45
Many studies (n = 9/13) incorporated quantitative forms of assessments (e.g. 3DGA, weightbearing radiographs, electromyography (EMG), pressure switches) in an attempt to improve objectivity of the grading within the clinical tools.4,5,9,13,17,29,37-39 Improvements were reported in weightbearing radiological measurements of the hindfoot, midfoot, and forefoot,41 and some improvements in ankle kinematics noted in two studies using gait analysis.36,39 However, there was a lack of standardized approaches for conducting (and quantifying) the gait assessments to minimize accuracy errors, and few studies provided sufficient detail regarding who graded the assessments, raising the risk of observer bias. Despite these limitations, the improved radiological findings and the changed foot (ankle joint) biomechanics supported the high rates of success recorded by the clinical grading tools. For future studies, pre- and postoperative weightbearing radiographs and dynamic measures such as EMG,43 dynamic pedobarography,46 and 3-DGA47 (using multisegmented foot models) are recommended. Some studies described gross motor function either by labelling participants as ‘community walkers’,10,11 or by reporting pre-/postoperative GMFCS levels.8,14 Though it should be noted by the reader that the GMFCS is not recommended for use as an outcome measure, both studies reporting on postoperative classifications reported a tendency towards favourable changes in GMFCS levels. While suggestive of improvements in functional mobility, these approaches lack the precision needed to quantify change within the possibly wide range of abilities, and use classification systems in ways for which they were not designed. Measuring specific gait activity-related capabilities such as the Functional Mobility Scale,48 Timed Up and Go, or the Gross Motor Function Measure (GMFM), would facilitate measures of functional gait changes.
The age of the child at the time of intervention and the severity of the foot deformity appear intertwined and careful consideration of these factors should be evaluated further. Although data were unable to be presented or subanalyzed within this review by age of receipt of the surgery, the mean age at time of the surgery was nine years (3.3 to 26). Barnes and Herring11 reported two participants, both over the age of ten years at the time of the intervention with poor outcomes, and both of whom had fixed foot deformities preoperatively with an element of fixed varus of the hindfoot. Similarly, Scott and Scarborough17 identified three poor ratings (out of 17 feet) in the varus comparison group, associating these outcomes with older age at time of surgery (10.5, 13.3, and 17.2 years). Previous discussions on the optimum age for intervention advise that the child is older than six years of age,9,13 as it is thought that rapid bone growth (< 6 years) can affect the stability of the attachment over time and increase risk of recurrences.12 It has also been suggested that there is a higher risk of subsequent valgus deformity when surgery is performed under the age of eight years.9,19 On the other hand, studies in this review have reported varus recurrences in patients older than ten years.9,17 Wong et al’s9 cohort reported one recurrent varus (10 years at index) and two persistent varus (15 and 16 years at index) deformities. The cohort in the Vlachou et al13 study had an average of 12.4 years at index operation and reported four recurrent varus deformities. These reports are consistent with the literature, advocating for surgery to be implemented in the early stages of deformity prior to bone maturity and the development of ‘fixed’ deformities.49,50 In addition to minimizing fixed deformities in the older years, there is widespread agreement in the studies reviewed that early (i.e. age < 6 years) surgery was a major risk factor for recurrent foot deformity. Therefore, a defined window of opportunity may be required for optimal results, and future research would be beneficial to assess this further.
Recurrence of foot deformity is an issue for discussion. For optimum results following SPLATT and SPOTT interventions, a number of authors have proposed that the patient should have ‘a dynamic passively’ correctable foot as a preoperative criterion.11,38 However, foot deformity is not a binary grading, rather a gradual transition within a spectrum. Patterns of muscle spasticity, tightness, and/or dystonia can also evolve over time, which may cause recurrence, or even reversal, of deformity. Exploring possible preoperative characteristics of participants with lower success rates may aid in the planning of and screening for patient suitability, as previously indicated by Chang et al.19 We were not able to identify features that indicated a better outcome from this review. However, previous papers have reported that children with hemiplegic CP,13,14,40 and lower GMFCS levels, show trends towards higher success rates.46 Nearly all the studies included some participants with tendoachilles lengthening; future studies may benefit from considering the combination of achilles lengthening and split tibialis tendon transfers. A better understanding of the causes of relapse or other foot deformity may come from pooling of multisite data, with standardized data collection prior and post-surgery.
The finding and interpretation of this review are limited by the availability of participant data (isolated to the target intervention), and the heterogeneity of reporting and selected outcome measures. Attempts made to contact authors proved particularly difficult in studies published over a decade prior (resulting in exclusion (Supplementary Table i)).34,45,49,51,52 Details regarding preoperative levels of ambulation and the degree of varus or EQV foot deformity were at times unclear, and in some early studies the underlying assumption has been made that all participants described as having CP actually meet the current definition of CP.5,10,38 While we have attempted to discuss outcomes in relation to their preoperative status of deformity, the transition from flexible to stiff foot is slow, gradual, and not binary. Pre- and postoperative pre/rehabilitation programmes may also impact patient outcomes, but details on these protocols were sparse (i.e. rate, period, patient adherence). Supplementary data in this domain could be beneficial, adding value to future clinical practice.
In summary, split tibial tendon transfers for the management of EQV foot deformities appear to be an effective intervention, producing and maintaining a plantigrade functional foot, eliminating the need for orthotics during gait and improving shoe wear in children and youth with CP.7,40 Post-surgery, approximately 87% of patients were able to walk with a plantigrade balanced foot without the need for an orthosis, reducing the risks of trips and falls and thus enhancing gait function.34 However, much of the evidence for this surgery comes from studies dating prior to 2011, with low-quality ratings and moderate risk of bias, favouring the use of Kling’s observational criteria to evaluate outcomes. Nevertheless, the data presented can be used to guide prospective study designs. The recurrence of deformity and occurrence of other foot deformities necessitating revision surgery is significant, and future work needs to focus on reducing the rate of revision surgery through better patient selection and more targeted surgery.
References
1. Sees JP , Miller F . Overview of foot deformity management in children with cerebral palsy . J Child Orthop . 2013 ; 7 ( 5 ): 373 – 377 . Crossref , PubMed Google Scholar
2. Krzak JJ , Corcos DM , Damiano DL , et al. Kinematic foot types in youth with equinovarus secondary to hemiplegia . Gait Posture . 2015 ; 41 ( 2 ): 402 – 408 . Crossref , PubMed Google Scholar
3. Gourdine-Shaw MC , Lamm BM , Herzenberg JE , Bhave A . Equinus deformity in the pediatric patient: causes, evaluation, and management . Clin Podiatr Med Surg . 2010 ; 27 ( 1 ): 25 – 42 . Crossref , PubMed Google Scholar
4. Green NE , Griffin PP , Shiavi R . Split posterior tibial-tendon transfer in spastic cerebral palsy . J Bone Joint Surg Am . 1983 ; 65 ( 6 ): 748 – 754 . PubMed Google Scholar
5. Kling TF , Kaufer H , Hensinger RN . Split posterior tibial-tendon transfers in children with cerebral spastic paralysis and equinovarus deformity . J Bone Joint Surg Am . 1985 ; 67 ( 2 ): 186 – 194 . PubMed Google Scholar
6. Lee CL , Bleck EE . Surgical correction of equinus deformity in cerebral palsy . Dev Med Child Neurol . 1980 ; 22 ( 3 ): 287 – 292 . Crossref , PubMed Google Scholar
7. Lullo B , Nazareth A , Rethlefsen S , Illingworth KD , Abousamra O , Kay RM . Split Tibialis Anterior Tendon Transfer to The Peroneus Brevis or Tertius for the Treatment of Varus Foot Deformities in Children with Static Encephalopathy: A retrospective case series . J Am Acad Orthop Surg Glob Res Rev . 2020 ; 4 ( 5 ): e2000044 . Crossref , PubMed Google Scholar
8. Sarıkaya İA , Birsel SE , Şeker A , Erdal OA , Görgün B , İnan M . The split transfer of tibialis anterior tendon to peroneus tertius tendon for equinovarus foot in children with cerebral palsy . Acta Orthop Traumatol Turc . 2020 ; 54 ( 3 ): 262 – 268 . Crossref , PubMed Google Scholar
9. Wong P , Fransch S , Gallagher C , et al. Split anterior tibialis tendon transfer to peroneus brevis for spastic equinovarus in children with hemiplegia . J Child Orthop . 2021 ; 15 ( 3 ): 279 – 290 . Crossref , PubMed Google Scholar
10. Hoffer MM , Barakat G , Koffman M . 10-year follow-up of split anterior tibial tendon transfer in cerebral palsied patients with spastic equinovarus deformity . J Pediatr Orthop . 1985 ; 5 ( 4 ): 432 – 434 . Crossref , PubMed Google Scholar
11. Barnes MJ , Herring JA . Combined split anterior tibial-tendon transfer and intramuscular lengthening of the posterior tibial tendon. Results in patients who have a varus deformity of the foot due to spastic cerebral palsy . J Bone Joint Surg Am . 1991 ; 73 ( 5 ): 734 – 738 . PubMed Google Scholar
12. Ahmed GS , Shaikh BF , Memon AR . Surgical treatment of equinovarus deformity of foot in children with cerebral palsy . Medical Channel . 2011 ; 17 ( 3 ): 21 – 24 . PubMed Google Scholar
13. Vlachou M , Dimitriadis D . Split tendon transfers for the correction of spastic varus foot deformity: a case series study . J Foot Ankle Res . 2010 ; 3 : 28 . Crossref , PubMed Google Scholar
14. Limpaphayom N , Chantarasongsuk B , Osateerakun P , Prasongchin P . The split anterior tibialis tendon transfer procedure for spastic equinovarus foot in children with cerebral palsy: results and factors associated with a failed outcome . Int Orthop . 2015 ; 39 ( 8 ): 1593 – 1598 . Crossref , PubMed Google Scholar
15. Aleksić M , Baščarevic Z , Stevanović V , Rakočević J , Baljozović A , Čobeljić G . Modified split tendon transfer of posterior tibialis muscle in the treatment of spastic equinovarus foot deformity: long-term results and comparison with the standard procedure . Int Orthop . 2020 ; 44 ( 1 ): 155 – 160 . Crossref , PubMed Google Scholar
16. Michlitsch MG , Rethlefsen SA , Kay RM . The contributions of anterior and posterior tibialis dysfunction to varus foot deformity in patients with cerebral palsy . J Bone Joint Surg Am . 2006 ; 88 ( 8 ): 1764 – 1768 . Crossref , PubMed Google Scholar
17. Scott AC , Scarborough N . The use of dynamic EMG in predicting the outcome of split posterior tibial tendon transfers in spastic hemiplegia . J Pediatr Orthop . 2006 ; 26 ( 6 ): 777 – 780 . Crossref , PubMed Google Scholar
18. Rethlefsen SA , Healy BS , Wren TAL , Skaggs DL , Kay RM . Causes of intoeing gait in children with cerebral palsy . J Bone Joint Surg Am . 2006 ; 88 ( 10 ): 2175 – 2180 . Crossref , PubMed Google Scholar
19. Chang CH , Albarracin JP , Lipton GE , Miller F . Long-term follow-up of surgery for equinovarus foot deformity in children with cerebral palsy . J Pediatr Orthop . 2002 ; 22 ( 6 ): 792 – 799 . PubMed Google Scholar
20. Graham HK , Baker R , Dobson F , Morris ME . Multilevel orthopaedic surgery in group IV spastic hemiplegia . J Bone Joint Surg Br . 2005 ; 87-B ( 4 ): 548 – 555 . Crossref , PubMed Google Scholar
21. Root L , Miller SR , Kirz P . Posterior tibial-tendon transfer in patients with cerebral palsy . J Bone Joint Surg Am . 1987 ; 69 ( 8 ): 1133 – 1139 . Crossref , PubMed Google Scholar
22. Grzegorzewski A , Borowski A , Pruszczyński B , Wranicz A , Domzalski M , Synder M . Split tibialis posterior tendon transfer on peroneus brevis for equinovarus foot in CP children . Chir Narzadow Ruchu Ortop Pol . 2007 ; 72 ( 2 ): 117 – 120 . PubMed Google Scholar
23. Fucs P , Kertzman PF , Svartman C . Treatment of spastic cerebral palsied patient with varus foot by split tibial tendon transfer . Rev Bras Ortop . 1997 ; 32 ( 1 ): 17 – 20 . Google Scholar
24. No authors listed . Global Issues: Youth . United Nations . https://www.un.org/en/global-issues/youth#:~:text=There%20is%20no%20universally%20agreed,of%2015%20and%2024%20years ( date last accessed 18 April 2023 ). Google Scholar
25. Moher D , Shamseer L , Clarke M , et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement . Syst Rev . 2015 ; 4 ( 1 ): 1 . Crossref , PubMed Google Scholar
26. Vogt JC . Split anterior tibial transfer for spastic equinovarus foot deformity: retrospective study of 73 operated feet . J Foot Ankle Surg . 1998 ; 37 ( 1 ): 2 – 7 . Crossref , PubMed Google Scholar
27. Miller GM , Hsu JD , Hoffer MM , Rentfro R . Posterior tibial tendon transfer: a review of the literature and analysis of 74 procedures . J Pediatr Orthop . 1982 ; 2 ( 4 ): 363 – 370 . PubMed Google Scholar
28. Park CI , Park ES , Kim HW , Rha D-W . Soft tissue surgery for equinus deformity in spastic hemiplegic cerebral palsy: effects on kinematic and kinetic parameters . Yonsei Med J . 2006 ; 47 ( 5 ): 657 – 666 . Crossref , PubMed Google Scholar
29. Saji MJ , Upadhyay SS , Hsu LC , Leong JC . Split tibialis posterior transfer for equinovarus deformity in cerebral palsy. Long-term results of a new surgical procedure . J Bone Joint Surg Br . 1993 ; 75 ( 3 ): 498 – 501 . Crossref , PubMed Google Scholar
30. World Health Organization . World Health Organization International Classification of Functioning, Disibility and Health (ICF) . Geneva, Switzerland , 2001 . Google Scholar
31. Slim K , Nini E , Forestier D , Kwiatkowski F , Panis Y , Chipponi J . Methodological index for non-randomized studies (minors): development and validation of a new instrument . ANZ J Surg . 2003 ; 73 ( 9 ): 712 – 716 . Crossref , PubMed Google Scholar
32. Sterne JA , Hernán MA , Reeves BC , et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions . BMJ . 2016 ; 355 : i4919 . Crossref , PubMed Google Scholar
33. Guyatt G , Oxman AD , Akl EA , et al. GRADE guidelines: 1. Introduction-GRADE evidence profiles and summary of findings tables . J Clin Epidemiol . 2011 ; 64 ( 4 ): 383 – 394 . Crossref , PubMed Google Scholar
34. Bickley C , Linton J , Scarborough N , Sullivan E , Mitchell K , Barnes D . Correlation of technical surgical goals to the GDI and investigation of post-operative GDI change in children with cerebral palsy . Gait Posture . 2017 ; 55 : 121 – 125 . Crossref , PubMed Google Scholar
35. Vlachou M , Beris A , Dimitriadis D . Split tibialis posterior tendon transfer for correction of spastic equinovarus hindfoot deformity . Acta Orthop Belg . 2010 ; 76 ( 5 ): 651 – 657 . PubMed Google Scholar
36. Dussa CU , Böhm H , Döderlein L , Fujak A . Treatment of spastic varus/ equinovarus foot with split-tendon transfers in cerebral palsy: How does it affect the hindfoot motion? Gait Posture . 2022 ; 92 : 343 – 350 . Crossref , PubMed Google Scholar
37. Mulier T , Moens P , Molenaers G , Spaepen D , Dereymaeker G , Fabry G . Split posterior tibial tendon transfer through the interosseus membrane in spastic equinovarus deformity . Foot Ankle Int . 1995 ; 16 ( 12 ): 754 – 759 . Crossref , PubMed Google Scholar
38. Synder M , Kumar SJ , Stecyk MD . Split tibialis posterior tendon transfer and tendo-Achillis lengthening for spastic equinovarus feet . J Pediatr Orthop . 1993 ; 13 ( 1 ): 20 – 23 . PubMed Google Scholar
39. O’Byrne JM , Kennedy A , Jenkinson A , O’Brien TM . Split tibialis posterior tendon transfer in the treatment of spastic equinovarus foot . J Pediatr Orthop . 1997 ; 17 ( 4 ): 481 – 485 . PubMed Google Scholar
40. Aleksić M , Baščarevic Z , Stevanović V , Rakočević J , Baljozović A , Čobeljić G . Modified split tendon transfer of posterior tibialis muscle in the treatment of spastic equinovarus foot deformity: long-term results and comparison with the standard procedure . Int Orthop . 2020 ; 44 ( 1 ): 155 – 160 . Crossref , PubMed Google Scholar
41. Davids JR , Gibson TW , Pugh LI . Quantitative segmental analysis of weight-bearing radiographs of the foot and ankle for children: normal alignment . J Pediatr Orthop . 2005 ; 25 ( 6 ): 769 – 776 . Crossref , PubMed Google Scholar
42. Tabaie SA , Videckis AJ , Quan T , Sheppard ED . Topical Review: Approach to Diagnosis and Management of the Pediatric Foot and Ankle in Cerebral Palsy Patients . Foot Ankle Orthop . 2022 ; 7 ( 2 ): 24730114221091800 . Crossref , PubMed Google Scholar
43. Almoajil H , Hopewell S , Dawes H , Toye F , Theologis T . A core outcome set for lower limb orthopaedic surgery for children with cerebral palsy: An international multi-stakeholder consensus study . Dev Med Child Neurol . 2023 ; 65 ( 2 ): 254 – 263 . Crossref , PubMed Google Scholar
44. Thomason P , Tan A , Donnan A , Rodda J , Graham HK , Narayanan U . The Gait Outcomes Assessment List (GOAL): validation of a new assessment of gait function for children with cerebral palsy . Dev Med Child Neurol . 2018 ; 60 ( 6 ): 618 – 623 . Crossref , PubMed Google Scholar
45. Fucs P , Kertzman PF , Svartman C . Treatment of spastic cerebral palsied patient with varus foot by split tibial tendon transfer . Rev Bras Ortop . 1997 ; 32 ( 1 ): 17 – 20 . Google Scholar
46. Chang CH , Miller F , Schuyler J . Dynamic pedobarograph in evaluation of varus and valgus foot deformities . J Pediatr Orthop . 2002 ; 22 ( 6 ): 813 – 818 . PubMed Google Scholar
47. Stebbins J , Harrington M , Thompson N , Zavatsky A , Theologis T . Repeatability of a model for measuring multi-segment foot kinematics in children . Gait Posture . 2006 ; 23 ( 4 ): 401 – 410 . Crossref , PubMed Google Scholar
48. Harvey AR , Morris ME , Graham HK , Wolfe R , Baker R . Reliability of the functional mobility scale for children with cerebral palsy . Phys Occup Ther Pediatr . 2010 ; 30 ( 2 ): 139 – 149 . Crossref , PubMed Google Scholar
49. Grzegorzewski A , Borowski A , Pruszczynski B , Wranicz A , Domzalski M , Synder M . [Split tibialis posterior tendon transfer on peroneus brevis for equinovarus foot in CP children] . Chir Narzadow Ruchu Ortop Pol . 2007 ; 72 ( 2 ): 117 – 120 . [Article in Polish].PubMed Google Scholar
50. Ruda R , Frost HM . Cerebral palsy. Spastic varus and forefoot adductus, treated by intramuscular posterior tibial tendon lengthening . Clin Orthop Relat Res . 1971 ; 79 : 61 – 70 . PubMed Google Scholar
51. Root L . Varus and valgus foot in cerebral palsy and its management . Foot Ankle . 1984 ; 4 ( 4 ): 174 – 179 . Crossref , PubMed Google Scholar
52. Romanini L , Villani C , Amorese V . Transfer of the posterior tibial tendon for equino-varus foot in cerebral palsy . Chirurgia del Piede . 1985 ; 9 ( 1 ): 7 – 15 . Google Scholar
Author contributions
J. Hochstetter-Owen: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing.
S. Stott: Conceptualization, Investigation, Writing – review & editing.
S. A. Williams: Conceptualization, Data curation, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing.
Funding statement
The authors received no financial or material support for the research, authorship, and/or publication of this article.
ICMJE COI statement
No external funding was received in support of this study and the authors have stated that they had no interest that could be perceived as posing a conflict of bias. S. Stott is employed by the Auckland District Health Board and will claim back reimbursement of the processing fee for this article as part of their continuing medical education allowance. S. Stott also reports research funding from Auckland Medical Research Foundation and support from Health Research Council of New Zealand (unrelated to this study), and is a voluntary Chair of the Cerebral Palsy Clinical Network, Paediatric Society of New Zealand. S. A. Williams reports research funding from Auckland Medical Research Foundation, Health Research Council of New Zealand, and the Cerebral Palsy Alliance Research Foundation, all of which is unrelated to this study. S. A. Williams is also a voluntary board member of the Australasian Academy of Cerebral Palsy and Developmental Medicine, and a member of the governing council for the International Alliance of Academies of Childhood Disability. S. A. Williams's salary is supported by the Aotearoa Fellowship from the Robertson Foundation.
Data sharing
The data that support the findings for this study are available to other researchers from the corresponding author upon reasonable request
Acknowledgements
We would like to thank Professor H. Kerr Graham, Professor Roslyn N. Boyd, Susan Rethlefsen, Chakravarthy U Dussa and Professor Noppachart Limpaphayom for their assistance. J. Hochstetter-Owen had guidance from NHMRC-Centre for Research Excellence, Clinical CP Trials Network centre for research excellence: 2021 systematic review course and GRADE training workshop.
Open access funding
The authors confirm that the open access fee for this study was self-funded (Department of Surgery, Faculty of Medical & Health Science, University of Auckland, Auckland, New Zealand).
Follow S. A. Williams @WilliamsSianA
Supplementary material
The supplementary material includes an example of the MESH search strategies, a table outlining the scoring for the MINORS for each study, the ROBINS-I outcomes, GRADE tables, and a summary of reasons for exclusion for full-text articles.
© 2023 Author(s) et al. This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives (CC BY-NC-ND 4.0) licence, which permits the copying and redistribution of the work only, and provided the original author and source are credited. See https://creativecommons.org/licenses/by-nc-nd/4.0/
