Selective peripheral neurotomy (SPN) for spasticity in childhood

SPECIAL ANNUAL ISSUE Selective peripheral neurotomy (SPN) for spasticity in childhood M. P. Sindou & F. Simon & P. Mertens & P. Decq Received: 24 March 2007 /Published online: 29 June 2007 # Springer-Verlag 2007 Abstract Overview Excess spasticity leads to disability that is marked by impaired locomotion, handicapping deformities and, if not controlled, discomfort and pain. Selective peripheral neuro- tomy in the child is indicated for severe focal spasticity, when botulinum toxin injections cannot delay surgery any longer. Materials and methods Preoperative motor blocks mimick- ing the outcome of the surgical procedure are essential to establish the objectives of neurotomy. In the lower limb, obturator neurotomy is indicated for spasticity in the adductor muscles, hamstring neurotomy for the knee flexion and tibial neurotomy for the spastic foot. Anterior tibial neurotomy is indicated for the extensor hallucis spasticity and femoral neurotomy for spasticity in the quadriceps. In the upper limb, neurotomy of the pectoralis major and teres major nerves is indicated for spasticity of the internal rotators of the shoulder. Neurotomy of the musculocutaneous nerve is indicated for spasticity of the flexors of the elbow, and neurotomy of median and ulnar nerves are indicated for spasticity of the pronators and flexors of the wrist and fingers. Conclusion Selective peripheral neurotomy is a valuable neurosurgical procedure in well-trained surgical hands for severe focalised spasticity. Keywords Spasticity . Child . Ablative neurosurgical procedure . Functional neurosurgery . Selective neurotomy . Peripheral nerve Introduction Spasticity in the child should not be treated just because it is present, as it may be useful for compensating for the loss of motor power. Spasticity should only be treatedwhen excess tone leads to further functional losses, impairs locomotion or induces deformities. Functional neurosurgery should be considered when harmful spasticity cannot be controlled by physical therapy, medications and botulinum toxin. The goal of this article is to summarise the state of art of selective peripheral neurotomies (SPN) for disabling spasticity in children. The first part of the article reviews the mechanisms of action of SPN, assessment of spasticity in paediatric patients and technical principles of SPN. Next, the operative techniques for the lower and then the upper limb are described. Last, indications for SPN are presented. SPN is indicated for severe, focal spasticity in one (or a few) muscular group(s) in a limb. As features of disabling spasticity differ from one child to another, the general rule is to define and tailor within a multidisciplinary team the appropriate neurosurgical program for each candidate. Principles Rationale 1 Surgery should be performed so that excessive hyperto- nia is reduced without the suppression of useful muscular tone or impairment of residual motor and sensory functions (Fig. 1). Indeed, SPN aims at re-equilibrating the tonic Childs Nerv Syst (2007) 23:957–970 DOI 10.1007/s00381-007-0399-1 M. P. Sindou (*) : F. Simon : P. Mertens Department of Neurosurgery, Hôpital Neurologique Pierre Wertheimer, Université Claude-Bernard Lyon I, 59 Bd Pinel, 69003 Lyon, France e-mail: [email protected] P. Decq Service de Neurochirurgie, Hôpital universitaire Henri Mondor, 94010 Créteil cedex, France balance between agonist and antagonist muscles by reducing excess spasticity [5, 8, 16]. Decreased spasticity is obtained by sectioning both afferents and efferent fibres of the stretch–reflex at the level of a muscle’s nerve [30]. Thanks to fine microsurgical dissection and mapping with intra-operative electrical nerve stimulation [17], SPN can, in effect, be selective and effective [3, 22, 23]. 2 Neurotomy consists in partial sectioning of one or several motor branches of the nerves innervating targeted muscle(s) in which spasticity is considered to be excessive. Therefore, neurotomy must never involve a nerve trunk composed of a mixture of motor and cutaneous sensory nerve fibres, as even a partial section of these fibres can be responsible for deaf- ferentation pain. The motor branches must be either already clearly isolated from the nerve trunk or dissected and identified within the fascicles of the nerve trunk, several centimetres proximal to the formation of an identifiable branch. Several concerns must be considered. Some afferent fibres to the spinal cord participate in various segmental reflexes, and sectioning them eliminates spasticity [13, 26]. However, mild spasticity can persist in certain circum- stances such as when a child becomes anxious or when unusual effort is being extended to complete a motor task. In such cases, the spasticity is mediated by the few remaining intact afferent fibres. Some motoneuron axons are sectioned, thereby inducing motor paresis proportional to the number of axons sectioned. There is no scientific basis for defining the extent of this partial section. However, all surgeons agree that a partial neurotomy must include the sectioning of 50–80% (usually 75%) of all branches to a targeted muscle in order for it to be effective. 3 Neurotomy therefore interrupts segmental reflex arch by acting on both afferent and efferent pathways. Neurotomy eliminates the afferent pathway corresponding to proprio- ception of the muscle concerned and induces paralysis by section of the efferent pathway. The principle of neurotomy is based on the different plasticity of these two pathways. The nerve endings of intact α motoneurons sprout in an attempt to compensate for the partial denervation as re- growth has been prevented by the resection performed during neurotomy. New motor endplates appear, and the size of motor units increases in proportion to the degree of denervation [12]. On average, motor units increase by about five times their original size, allowing them to compensate for a loss of as much as 80% of the motoneuron pool [2]. In animal studies [24], sprouting zones have appeared 2 days after partial section of motor nerves [19]. The time to onset of sprouting in humans cannot be determined precisely, but it appears to be short. In a study of the long-term course of the H-reflex after tibial neurotomy, sprouting, as evaluated by return of the amplitude of the Mmax response to its baseline value, occurred 8 months after the operation at the latest. Re-innervation of sensory spindle nerve endings occurs after nerve section. However, this re-innervation is disor- ganised: Ib fibres can colonise primary or secondary spindle nerve endings and then become sensitive to stretching. This non-specific re-innervation of muscle receptors is largely non-functional: Study of the long-term course of the Hmax/Mmax ratio after tibial neurotomy has shown the absence of functional sensory re-innervation, as reflected by the absence of long-term recovery of the H response. Fig. 1 Methods for managing spasticity based on whether it is focal or general and on whether it is permanent or temporary 958 Childs Nerv Syst (2007) 23:957–970 Preoperative assessment Therapeutic objective The clinical examination should not be limited to the analytic assessment of spasticity. It must consider the child’s general state (e.g. hypotrophy, respiratory restrictive syndrome, swal- lowing difficulties), orthopaedic status (musculoskeletal con- tractures), evolution of motor function as measured with the Gross Motor FunctionMeasure (GMFM; decrease, increase or stable) [18] and cognitive and psychological abilities (able to participate in rehabilitation after surgery). The objective is to distinguish useful spasticity (e.g. the potential to stand for transfers) from harmful spasticity (e.g. standing on tip-toes). Clinical examination The clinical examination should include observation of posture and gait, assessment of passive range of motion, quantification of the degree of spasticity using Ashworth and Tardieu scales [31] and functional quantification using GMFM [28]. (See Appendix 2). Motor blocks Before recommending SPN, a test using nerve blocks of motor nerves innervating the targeted muscle is of prime importance [23]. These blocks, using local anaesthetics, such as long-lasting bupivacaine, enable the surgeon to evaluate the motor strength of antagonist muscles and determine if articular motion limitations are due to spasticity, muscu- lotendinous contractures or articular ankyloses. This test evaluates the proportion of spasticity mediated by the infiltrated nerve that is resulting in the limited range of movement of the joint concerned. In addition, botulinum toxin injections can be used as a ‘prolonged’ test for several weeks or months before pursuing neurosurgical treatment, as their effects will mimic the outcome of selective neurotomies on the injected nerves. This strategy of using pre-operative injection tests allows the child to appreciate the benefit that can follow a selective neurotomy [14]. Consequently, the objectives of treatment are easier to establish. Objectives may be cosmetic (e.g. to allow the child to put his or her hand in a pocket), related to nursing (e.g. to allow a caregiver to wash the palm of the child’s hand) or functional. Technical basis of neurotomy Anaesthesia Neurotomy is performed under general anaesthesia but without long-lasting muscle paralysis. Patient positioning and anaesthesia induction is not always easy because passive mobility is often limited. Most importantly, it may be useful to test the efficacy of the anaesthesia intra- operatively by evaluating the stretch reflex (e.g. test for clonus), as its presence implies that reflex pathways are not depressed by the anaesthetic drugs. Changes in certain reflexes may need to be recorded (such as the H reflex) to guide the operation. Therefore, the general anaesthesia has to be performed without long-lasting muscle relaxants so that the motor responses elicited by bipolar electrical stimulation of motor branches can be detected to identify the nerve. In addition, nitrogen monoxide and propofol are contra-indicated because they modify reflex excitability. Mapping Mapping is the anatomical identification of motor branches. It is an essential step to avoid sensory impairment and usually requires the use of the operating microscope. Frequent variation in the emergence of nerve branches and limited surgical access can nevertheless make this a difficult step. Consequently, a good quality intra-operative stimulation device must be available. The identification is based on descriptive anatomy but also on the muscular responses to bipolar electrical stimulation (e.g. NIMBUS®: Multifunctional Stimulator, Newmedics France). Stimula- tion is performed at low intensity ordinarily (1 mA) to avoid electrical diffusion and an incorrect interpretation. Usually, a triple-stimulation hook, composed of an anode between two cathodes, is used to grasp the nerve branch to be stimulated. Finally, the response to stimulation must be visualised in the form of movement of the limb segment. Section Once all of motor branches have been identified by electrical stimulation, those considered responsible for harmful spastic- ity are marked separately with colored tapes. According to the pre-operative evaluation and subsequent programme, variable proportions (50–80% depending on the degree of preoperative spasticity) of the isolated motor branches of fascicles are resected under the operating microscope near the muscle to ensure that only its muscular branches are cut. The resection is 5 mm long from the proximal stump, which is coagulated with bipolar forceps to prevent re-growth of fibres. When there are several nerve branches for one muscle, one or more branches could easily be sectioned completely until the global amount needed for the considered muscle is attained. The effect of each nerve resection on spasticity is then evaluated by comparing muscle responses to electrical stimulation, proximally and then distally to the resected portion of the nerve. If the response after proximal stimulation is still intense, further resection can be Childs Nerv Syst (2007) 23:957–970 959 performed. The aim is to decrease motor innervation enough to avoid recurrence of spasticity by ‘take-over’ (re-innervation or ‘adoption’ of muscular fibres denervated as a result of the neurotomy by the surrounding intact motor nerve fibres). Operative techniques Surgery for lower limb Obturator neurotomy for the hip Obturator neurotomy (Fig. 2) eliminates spasticity in the adductor muscles. It is often recommended for diplegic xchildren with cerebral palsy when their walking is hampered by crossing of the lower limbs. It can also be used for paraplegic children to facilitate perineal toileting and self-catheterisation. The incision can be performed along the body of the adductor longus in the proximal part of the thigh (Fig. 2a-1). Over recent years, transverse incision in the hip flexion fold, centred on the prominence of the adductor longus tendon, has been preferred. In addition to its more aesthetic appearance, this incision facilitates adductor longus tenotomy when necessary (Fig. 2a-2). To rapidly locate the anterior branch of the obturator nerve, the dissection is conducted laterally to the adductor longus muscle body. The posterior branch is situated more deeply and should be spared to preserve the hip’s stabilizing muscles [1, 21] (Fig. 2b). Hamstring neurotomy for the knee Hamstring neurotomy (Fig. 3) is indicated in children with spastic diplegia to counter the accentuation of the flexion deformity of the knees observed with growth. The transverse incision is performed in the gluteal fold, centered on the groove between the ischium and the greater trochanter (Fig. 3a). After crossing the fibres of the gluteus maximus, the sciatic nerve is identified in the depth of the incision. The branches to the hamstring muscles are isolated at the border of the nerve, primarily based on responses of the semitendinosus muscle, which is often the major muscle responsible for spasticity [6] (Fig. 3b). Tibial neurotomy for the foot Tibial neurotomy is indicated for the treatment of spastic varus plantar flexion (equinus deformity) with or without claw toes [29]. It consists of exposing all motor branches of the tibial nerve at the popliteal fossa (i.e. the nerves to gastrocnemius and soleus, tibialis posterioris, popliteus, flexor hallucis longus and flexor digitorum longus). The soleus has been demonstrated to be usually almost fully Fig. 2 a Skin incision for (right) obturator neurotomy, on the relief of adductor longus muscle (1) or in the hip flexion fold centered on the prominence of the adductor longus tendon (2), which gives a cosmetic advantage. b Dissection of the anterior branch (AB) of the right obturator nerve (ON). The ad- ductor longus (AL) is retracted laterally and the gracilis (G) medially. The nerve is anterior to the adductor brevis (AB). Adductor brevis nerve (1), (2), adductor longus nerve (3), gra- cilis nerve (4), (5). The posterior branch (PB) of the obturator nerve lies under the adductor brevis (AB) and should be spared 960 Childs Nerv Syst (2007) 23:957–970 responsible for the pathogenesis of spastic plantar flexion allowing sparing of gastrocnemius [9]. The incision can be vertical, on either side of the popliteal fossa, and extended inferiorly. Over recent years, a transverse incision in the popliteal fossa, which gives a much better long-term aesthetic result, has been adopted. In addition, this transverse incision allows for a high of the gastrocnemius insertion fascia, if necessary, at the end of the operation, (Fig. 4a). The first nerve encountered is the sensory medial cutaneous nerve of the leg. Situated immediately anteriorly to the saphenous vein, it must be spared. More deeply, the tibial nerve trunk, from which the nerves to the gastrocne- mius emerge, is easily identified. The superior soleus nerve is situated in the midline, just posterior to the tibial nerve. The effect of a soleus neurotomy is assessed by the immediate intra-operative disappearance of the ankle clonus. By retracting the tibial nerve trunk medially using a traction suture, the other branches can be identified by electrical stimulation as they emerge from the lateral edge of the tibial nerve trunk. The most lateral branch is often the popliteal nerve, followed by the tibialis posterior nerve and finally by the inferior soleus and flexor digitorum longus nerves. Some fascicles, often larger, can give a toe flexion response via intrinsic toe flexors (Fig. 4b). However, neurotomy of these branches is not recommended if they cannot be clearly individualised at this level because they frequently are mixed with sensory fascicles [4, 11]. Anterior tibial neurotomy for extensor hallucis This neurotomy (rarely performed) is indicated for the treatment of permanent extension of the hallux when it is difficult to wear shoes and botulinum toxin injections have failed to manage the condition. In practice, this neurotomy may be indicated after unjustified section of the flexor hallucis tendon resulting in a disequilibrium that favours the extensor. A vertical incision is centered on the junction between the tibialis anterior and the extensor hallucis, at the middle third of the leg. The tibial nerve is situated deeply between these two muscle heads, and the neurotomy is performed on the motor branch to the extensor hallucis [4]. Femoral neurotomy for the quadriceps Femoral neurotomy is indicated to treat excessive spasticity of the quadriceps muscle. This muscle is very often spastic and can interfere with gait by limiting knee flexion during the swing phase. Given its ‘strategic’ importance in maintaining upright posture, a motor block is an essential part of the pre-operative evaluation. The neurotomy mainly concerns the motor branch to the rectus femoris and vastus intermedius muscles. The incision is horizontal in the hip flexion fold (Fig. 5a). The dissection passes medial to the sartorius muscle body and exposes the motor branches of the femoral nerve, first the nerve to the rectus femoris and then, more deeply, the nerve to the vastus intermedius (Fig. 5b). Fig. 3 a The skin incision for hamstring neurotomies (on the right side) is located on the midline between the ischial tu- berosity (IT) and the greater trochanter (GT) (1). A transverse incision can also be performed in the gluteal fold (2) centred on the groove between the ischial tuberosity and the greater tro- chanter, for better aesthetic results. b Dissection of the right sciatic nerve (SN), under the piriformis muscle (P), after passing through the fibres of the gluteus maximus muscle (GM). The epineurium of the nerve is opened, and fascicles for ham- string muscles (HF) are local- ised in the medial part of the nerve trunk. Inferior gluteal nerve (IGN), inferior gluteal artery (IGA) Childs Nerv Syst (2007) 23:957–970 961 Electrical stimulation is essential given the large number of sensory fascicles in this nerve that must be spared. Surgery for the upper limb Pectoralis major neurotomy for the shoulder Neurotomy of collateral branches of the brachial plexus innervating the pectoralis major is indicated for spasticity of the shoulder with internal rotation and adduction. The skin incision is made at the innermost part of the deltopectoral sulcus and curves along the clavicular axis. First, the clavipectoralis fascia is opened. Then, the upper border of the pectoralis major muscle is reflected downwards. Close to the thoracoacromialis artery, the ansa of the pectoralis muscle is identified with the aid of a nerve stimulator [7]. Teres major neurotomy for the shoulder Neurotomy of collateral branches of the brachial plexus innervating the teres major is also indicated for spasticity of the shoulder with internal rotation and adduction. The skin incision follows the inner border of the teres major, from the lower border of the deltoid muscle’s posterior head to the lower extremity of the scapula. The lower border of the long portion of brachii triceps constitutes the upper limit of the approach. The dissection continues deeply between the teres minor and major muscles. In the vicinity of the subscapularis artery, the nerve ending on the teres major is identified. The nerve is surrounded by thick fat when approaching the anterior facet of the muscle body [7]. Musculo-cutaneous neurotomy for the elbow Neurotomy of the musculo-cutaneous nerve is indicated for spasticity of the elbow with flexion, depending on the biceps brachii and the brachialis muscles. The skin incision is performed longitudinally. It extends from the inferior edge of the pectoralis major, medial to the biceps brachii, down to 5 cm (Fig. 6a). The superficial fascia is opened between the biceps laterally and the brachialis medially. The brachial artery and median nerve exit medially. The dissection proceeds in this space, where the musculo-cutaneous nerve lies anterior to the brachialis muscle (Fig. 6b). Opening the epineurium allows the fascicles of the nerve to be dissected under high magnification of the operating microscope. The motor fascicles are distinguished from the sensory ones using the nerve stimulator [15, 25]. Median neurotomy for the wrist and fingers Neurotomy of the median nerve is indicated for spasticity of the forearm with a pronation deformity resulting from spasticity in the pronator teres and quadratus muscles and a flexion deformity in the wrist because of spasticity in the flexor carpi radialis and palmaris longus muscles. In the hand, median neurotomy is indicated for flexion deformity in the fingers because of spasticity in the flexor digitorum superficialis (flexion of proximal interphalangeal joint and metacarpophalangeal joint) and flexor digitorum profondus (flexion of distal interphalangeal, proximal interphalangeal and metacarpophalangeal joint) muscles. Swan neck defor- mation of the fingers depending on the lumbrical and Fig. 4 a Skin incision of the tibial nerve, vertical incision either side of the right popliteal fossa, extending inferiorly (1) or transverse incision in the popliteal fossa (2) for a better long-term aesthetic result. b: Dissection of the tibial nerve, dorsal view of the right popliteal region.Tibial nerve (1), peroneal nerve (2). The sensory sural nerve (3) lies superficially just beneath the subcutaneous aponeurosis between the two gastrocnemius muscles. The medial and lateral gastrocnemius nerves (4) may arise either separately from the both sides of the tibial trunk or posteriorly from a common origin, sometimes including the sensory sural nerve. Each gastrocnemius nerve usually divides into two distal branches when approaching the muscle. The one or two soleus nerves (5) may arise from a common origin or quite separately from the tibial nerve. The posterior tibialis nerve (6), like the soleus nerve, originates from the ventro-lateral aspect of the tibial nerve but more distally at the level of the soleus arch. Sometimes, it may originate from a common trunk with the inferior branch of the soleus nerve. The distal trunk of the tibial nerve (7) contains five to eight fascicles averaging 1 mm in diameter each; two thirds of them are motor fascicles, and one third are sensory ones. Muscles: lateral and medial gastrocnemius (LG–MG), Soleus (S) 962 Childs Nerv Syst (2007) 23:957–970 interosseous muscles can be treated by neurotomy, these muscles being innervated by the median and ulnar nerves. Concerning the thumb, neurotomy of the median nerve is indicated for flexion and adduction/flexion deformities (thumb-in-palm deformity) because of spasticity in the flexor pollicis longus muscle. The skin incision begins 2 to 3 cm above the flexion line of the elbow, medial to the biceps brachii tendon, passes through the elbow and curves towards the junction of the upper and middle thirds of the anterior forearm (the convexity of the curve turns laterally; Fig. 5 a Skin incision for right femoral neurotomy, below the inguinal ligament, laterally to the femoral artery (1) or hori- zontal in the hip flexion fold (2) for better aesthetic results. b Dissection of the right femoral nerve (FN) and its branches, after opening the anterior fascia of the psoas muscle (P). The bipolar stimulation allows iden- tification of the two or three branches to sartorius muscle (S) and three or four to rectus femoris muscle, which produces flexion of the hip. The nerve to the vastus intermedius can be found more deeply. Femoral artery (FA), femoral vein (FV) Fig. 6 a Skin incision for right musculocutaneous neurotomy, along the medial aspect of the biceps brachii, under the inferior edge of the pectoralis major muscle. b Dissection of the right musculocutaneous nerve (MC) in the space between the biceps brachii (BB) laterally, the coraco brachialis (CB) medially and the brachialis (B) posteriorly. Branches to brachialis (1), (2) and to biceps brachii (3), (4) are recognised by stimulation giving elbow flexion. Humeral artery (H) with median nerve are situ- ated medially and are not dissected Childs Nerv Syst (2007) 23:957–970 963 Fig. 7a). Thereafter, the median nerve is searched medially to the brachial artery and recognised at the elbow, deeply under the lacertus fibrosus, which is cut. Sharp dissection is used to separate the branches of the median nerve. The pronator teres belly with its two heads is retracted medially and distally so that its muscular branches can be inspected. This muscle is retracted up and laterally while the flexor carpi radialis is pulled down and medially. The muscular branches to the flexor carpi radialis and to the flexor digitorum superficialis can then be seen. Finally, the latter is retracted medially uncovering the branches to the flexor digitorum profondus, the flexor pollicis longus and the pronator quadratus. These latter muscular branches may be individualised as separate branches or remain together in the distal trunk of the anterior inter-osseous nerve. Sometimes, it may be useful to divide the fibrous arch of the flexor digitorum superficialis muscle to make the dissection easier (Fig. 7b). Besides this approach, which requires a long incision and extensive dissection, a ‘minimal’ approach can be performed. The different fascicles in the trunk of the median nerve, just medial to the brachial artery, are dissected. This latter approach provides a better cosmetic outcome (shorter incision). However, it has the inconve- nience of offering a nerve exposure less propitious for identifying the various motor branches in the form of fascicles containing only fibres of the targeted muscle(s) owing to the fact that at this level of the nerve, cutaneous sensory fibres are mixed with the muscle fibres. This entails the risk of sensory complications, especially for developing a complex regional pain syndrome. Ulnar neurotomy for wrist and fingers Neurotomy of the ulnar nerve is also indicated for spastic wrist flexion and for spasticity in the flexor carpi ulnaris muscle. In the hand, ulnar neurotomy is indicated for flexion deformity in the fingers because of spasticity in the portion of the flexor digitorum profondus muscle (flexion of distal interphalangeal, proximal interphalangeal and metacarpophalangeal joint) innervated by the ulnar nerve. Ulnar neurotomy is also indicated for spasticity causing an adduction/flexion deformity of the thumb because of the adductor pollicis muscle and can be used in combination with median neurotomy to treat a swan-neck deformation of the fingers. A separate arched skin incision is performed to expose the ulnar nerve at the medial part of the elbow (Fig. 8). After subcutaneous dissection, the ulnar nerve is identified medially to the medial epicondyle, where it enters between the two heads of the flexor carpi ulnaris. There, the motor branches to the flexor carpi ulnaris muscle are identified. More distally, the branches to the medial half of the flexor digitorum profondus are found. Fig. 7 a Skin incision on the right forearm for median neurotomy from the medial aspect of the biceps brachii at the level of the elbow longitudinally along the bicipital crest (1). The incision can eventually be continued distally towards the midline above the wrist (2). b Dissection of the median nerve in two stages. First stage of the dissection (upper figure), the pronators teres (PT) is retracted upward and laterally and the flexor carpi radialis (FCR) medially. Branches from the median nerve (MN), before it passes under the fibrous arch of the flexor digitorum superficialis (FDS), are dissected: to the pronators teres (1) and two nerve trunks to the flexor carpi radialis, palmaris longus and flexor digitorum superficialis (2), (3). Second stage of the dissection (lower figure), the fibrous arch of the FDS is sectioned to allow a more distal dissection of the median nerve. The FDS is retracted medially and branches from the median nerve are identified: (1) to the flexor pollicis longus (FPL); (2) to the flexor digitorum profondus (FDP); (3) the inter-osseous nerve and its proper branches to these muscles 964 Childs Nerv Syst (2007) 23:957–970 Complications, side effects and causes of recurrence Local complications include post-operative haematoma and infection. They are rather rare if preventive measures are respected. Sensory disorders, such as paresthesiae followed by transient deafferentation pain (2 months), can be observed if the section accidentally includes several sensory fas- cicles. This complication underscores the importance of precise stimulation. Patients rarely complain of decreased muscle strength after neurotomy. Muscle function is redundant, and no single muscle is solely responsible for the movement of a body segment, without the possibility of substitution by another muscle. Specifically, in surgery of the upper limb, complications include rare transitory hypesthesia of the anterior part of the forearm because the surgical approach can result in the cutting of subcutaneous sensory branches. Paresis of flexors of the elbow, wrist, fingers or both (because of excessive nerves sectioning) is rare, transient and responds to physical therapy. Recurrence of spasticity can be observed when the amount of sectioning is insufficient. In such cases, re- operation can be performed. General complications can also be found, as in all peripheral surgical procedures, but they are rare. Principal among them is scarring. Indications General indications Based on the evaluations, the indication for neurotomy can be defined by observing a number of principles: 1. Neurotomies are indicated for a localised spastic disorder, although a combination of several neuroto- mies can sometimes be considered in the case of more diffuse disorders. 2. The total absence of motor power in the antagonists allows for only a transient result, as residual spasticity, even when minimal, gradually returns the limb segment to its initial deformity because of the absence of opposition muscle power. In the case of complete absence of antagonists, neurotomy can only be indicated in combination with secondary tendon transfer surgery, to restore the deficient antagonists muscle power. 3. Excessive contractures of spastic muscles considerably limit the efficacy and therefore the value of neurotomy. In this case, an orthopaedic procedure combining tendon lengthening and/or transfer and/or arthrodesis would be preferable. 4. Neurotomy is only indicated when no other non- surgical treatment option is available and after the patient has completed a well-conducted rehabilitation programme. An interval of 1 year between onset of the spastic disorder and surgery is typically required. This surgical procedure cannot be performed in isolation. It must be integrated into a long-term rehabilitation programme that is acceptable to the patient over time. 5. Finally, although excessive spasticity is almost always harmful, it can sometimes be useful to the patient. Correction of spasticity can eliminate a function acquired by the patient as a result of adaptation to his or her handicap. For example, elbow flexion spasticity can be useful in patients with limited voluntary move- ments as it allows them to carry certain loads or to transiently unload the contra-lateral, uninvolved upper limb. This type of useful spasticity must be thoroughly looked for and understood before proposing neurotomy. Functional gain must take precedence over simple reduction in muscle tone. When spasticity is focal, botulinum toxin can be a good complement to physiotherapy and posture. Indeed, this therapy is sometimes sufficient; however, SPN is indicated if botulinum toxin fails [16]. Predictability of outcome The quality of outcome depends on the accuracy of the pre- operative assessment and on the precision of the surgical procedure [30]. The result should be more or less similar to that observed after the motor block. In contrast to infiltration techniques (which are also less precise), the results obtained by neurotomy persist after several years [10, 29]. Whereas the 1-year recurrence rate for spasticity after surgery on lower limb for adults is 1%, spasticity tends to recur more frequently in younger children [2]. This emphasises the importance of botulinum toxin injections, which enables neurosurgical treatment to be delayed until the child reaches a more appropriate age for selective neurotomy. To ensure stability of the result of an orthopae- dic or neurosurgical operation in children, it is essential to control for growth and its effects on the operated sector by Fig. 8 Skin incision on the right forearm for ulnar neurotomy, either a longitudinal incision posteriorly to the medial epicondyle and medially to the olecranon at the elbow (1) or a transverse medial incision in the wrist fold (2), depending on the location of the spastic muscle Childs Nerv Syst (2007) 23:957–970 965 careful prescription of nocturnal posture splints and a lighter daytime appliance (e.g. a dorsiflexor splint for the ankle to avoid equinus, ideally with a hinge to allow dorsiflexion). In addition, whereas increasing contractures and deformities can only be treated by orthopaedic surgery, neurosurgical treatment of spasticity beforehand can help improve the outcome of these operations. The outcome of SPN also intimately depends on the post- operative care after a successful surgery. The limb can be raised to avoid edema and mobilised as soon a possible, physical therapy beginning on the second post-operative day. An anti-thrombotic therapy is administered for 5 days after surgery, and the patient is discharged on the seventh post- operative day. Immediately thereafter, a programme of physical and occupational therapy is undertaken in a rehabilitation centre for at least 4 weeks and at an outpatient facility for 2 months. Conclusions Indications for the treatment of spasticity in children must consider the child’s overall interests with careful consider- ation of the environmental factors, the child’s general and orthopaedic status and the evolution of the child’s function. This approach requires a multidisciplinary team and the full participation of the child and his or her parents. In well-selected patients, SPN can yield worthwhile effects on refractory, focal spasticity and its consequences. SPN can unmask residual motor function of (antagonist) muscles, make passive movements easier, increase comfort in daily activity, decrease pain because of spasticity and improve the child’s cosmetic appearance. Orthopaedic surgery can augment the neurectomy when irreversible joint deformities and muscle contractures are present. Preliminary tests with an anaesthetic block are crucial to predict the outcome as they will mimic the effects of the proposed neurotomies. Positive effects from botulinum toxin injection are also predictive of the efficacy of SPN. Operative planning with a detailed clinical assessment of each spastic muscle is essential to determine the degree of nerve section, thus avoiding insufficient surgery or excessive loss of motor strength in the concerned muscles. Neurotomy performed distally on the muscular branches allows the selectivity of the procedure and avoids sensory complications. The essential problem raised by this type of surgery concerns the possibility of objective evaluation of the result obtained. Rehabilitation plays a fundamental role in the management of these patients and in its absence, the benefit of the proposed operation will remain limited [20, 22, 27]. In well-trained surgical hands, SPN is a safe and valuable neurosurgical procedure thanks to microsurgery and intra-operative mapping by electrical nerve stimulation. Appendix 1: The concept of focal spasticity Focal spasticity is based on two concepts. Firstly, the population of skeletal muscles is heterogeneous: Not all muscle sectors involve the same degree of spasticity in the same person; some sectors are always spastic while other sectors of the same limb are not. Indeed, most skeletal muscles are composed of a mixture, in variable proportions, of rapid, slow and intermediate fibres, thus conditioning their contractile characteristics. Muscles involved in limb movements tend to have a majority of rapid fibres (large and fatigable, mainly working by shortening) in contrast to postural muscles (with small, slow and resistant fibres, mainly working by lengthening), which provide slow and long-lasting contractions. Whereas the developed force of muscles with rapid fibres lasts 2–3 min, the contraction of slow fibres can remain constant for more than an hour. In spastic patients, the mechanical properties of muscles (viscosity and elasticity) are modified after the first year. Type II (rapid) fibres are transformed into type I (slow) fibres, and therefore, the duration of muscle contraction lengthens. The number of sarcomeres decreases, and viscosity increases, thus exaggerating muscle stiffness. Most spastic muscles are composed of slow fibres, which contract eccentrically (by lengthening) under con- ditions of natural movement. This is the case, in particular, for the soleus, semitendinosus, vastus intermedius of quadriceps, flexor carpi radialis and brachialis muscles. Secondly, slow muscle fibres are innervated by small motoneurons and consequently are activated first. This characteristic is another reason why this muscle population is the most excitable when spasticity is present. These motoneu- rons are integrated with the surrounding inter-neuronal popu- lation forming numerous reflex circuits specifically organised to achieve a specific objective, such as voluntary movement. When considering SPN, it is therefore essential to bear in mind that focal spasticity is unique to each limb segment, considering the muscle’s individual fibre composition and local stretch reflex circuit organisation (Fig. 9). Appendix 2: Preoperative assessment Exaggeration of the tonic stretch reflex is detected clinically as resistance to passive mobilisation of the joint. The intensity of this resistance is proportional to the rate of stretching. It disappears when stretching is stopped, as the limb segment tends to return to its initial position. This resistance is ‘elastic’ and contrasts to the ‘plastic’ resistance (where the limb remains in the position at which mobi- lisation was stopped; Fig. 10). In the quadriceps, the stretch reflex can be evaluated by the pendulum test. The patient is placed in the supine position, with the trunk and thigh resting on a hard surface. The leg, in 966 Childs Nerv Syst (2007) 23:957–970 full extension, is held by the examiner. The leg is suddenly dropped, and the angle of flexion at which the rate of flexion stops or declines, and the amplitude, frequency and number of oscillations are recorded with a goniometer. Spastic limbs have a smaller number and amplitude of oscillation, which can be described mathematically. Stretching a limb segment can trigger clonus. It is always associated with exaggerated deep tendon reflexes, but the opposite is not always true. Clonus can be either exhaust- ible or inexhaustible. In the lower limb, it is often observed at the ankle and sometimes in the quadriceps muscle. It can interfere with standing or gait. In the upper limb, clonus can interfere with the stability of standing and the residual function of the affected limb. It is always poorly tolerated in the upper limb, as it is perceived as an abnormal movement in addition to other handicapping features. Exaggeration of the stretch reflex is evaluated by the Ashworth score (Table 1), which, despite many imperfec- tions and criticisms, is widely used for the evaluation of spasticity. The intra-observer variability is satisfactory, but the inter-observer variability is not. The Tardieu scale (Table 2), however, integrates the concept of velocity- related variations and the concept of clonus. The inter- observer variability is better than that of Ashworth’s scale. Fig. 9 Decision algorithm for neurosurgical treatment of lower limb spasticity in children Childs Nerv Syst (2007) 23:957–970 967 Clinical examination involves the following: 1. First stage: observation of posture and gait The observation of the child in a lying position permits identification of asymmetric muscle tone, which is evidence of predominant hypertonia in particular muscular groups. Observation is particularly important for non-ambulatory children. They most often exhibit left or right windswept deformities or bilateral flexion–internal rotation–adduction of lower limbs. Based on the clinical examination of gait, gait can be classified into five groups: true equinus, jump gait, apparent equinus, crouch gait and asymmetric gait. These patterns reflect abnormal pelvic tilt, hip extension or flexion, knee extension or flexion, and ankle dorsi or plantar flexion. Each gait pattern is due to hypertonia in particular muscle groups, and they are identified for the management of their spasticity, contracture or both. This first clinical observation provides an idea of which muscular groups are weak and for which it would be Fig. 10 Decision algorithm for neurosurgical treatment of spas- ticity in children. General principles 968 Childs Nerv Syst (2007) 23:957–970 dangerous to decrease tonus. It shows which muscular groups have spasticity that is disturbing function and will be the target of treatment. It will also show what muscles are so contracted that they are inaccessible to neurosurgical treatment. 2. Second stage: assessment of range of motion When strength and tonus of some muscular groups deform the articulations into an asymmetric configuration, the range of motion of articulations decreases, and musculoskeletal contractures appear. It can then be too late to treat spasticity as a first intention. Orthopaedic surgery (lengthening contracted muscles, tendon transfer or osteo- tomy), with or without treatment for spasticity, must then be considered. Time is needed to measure range of motion. A gentle, slow approach is needed to avoid interference by hyperto- nicity and to be able to differentiate the implications of the spasticity and contractures that is present. 3. Third stage: quantification of the degree of spasticity The evaluation is performed while the patient is lying, rapidly moving the segments of the limbs: flexion extension of the knees, abduction of the hips, ankle dorsiflexion with knee extension and knee flexion. Both Ashworth and Tardieu scales are useful. For the assessment of spasticity, only one scale is insufficient to respond to all cases observed. The Tardieu scale seems to be appropriate for young children who walk and have no contractures. The Ashworth scale is best for the assessment of global spasticity in non-ambulatory children who are beginning to develop orthopaedic complications. 4. Fourth stage: functional quantification through GMFM Assessment of spasticity is just one element of global function measure. The GMFM is a clinical measure designed to evaluate changes in gross motor function in children with cerebral palsy. There are two versions of the GMFM: the original 88-item measure and the more recent 66-item GMFM. Items on the GMFM-88 span the spectrum of activities from lying and rolling to walking, running and jumping skills. The GMFM-66 includes a subset of the 88 items that has been shown to be dimensional. The GMFM-66 is the conversion of the ordinal scale into interval scale. This transformation increases the accuracy of a child’s ability and provides a measure that is equally responsive to change across the spectrum of ability. An assessment with GMFM every 6 or 12 months provides the curve of evolution and permits the motor potential of the child to be judged objectively. The findings can be extrapolated to future motor function or to observe a patient’s progress or degradation. The projected treatment can be based on a realistic perspective, and we can determine if a treatment can be delayed (if the curve shows increased function) or if urgent surgical treatment is needed (if the curve plateaus or decreases). Appendix 3: Electrophysiology The diagnosis of spasticity is clinical. However, it some- times can be useful to complete the clinical evaluation by electrophysiological recordings, especially to quantify the effect of treatments. Several electrophysiological methods are available to study the stretch reflex. The H reflex is usually recorded after a voluntary contraction, to pre-excite spinal motoneurons; however, it can be measured on a muscle in spastic activity even at rest. It contrasts with the T reflex (response to tendon percus- sion) because it is obtained by direct stimulation of proprioceptive fibres, whereas the latter, on the contrary, depends on the gain of the primary neuromuscular spindle nerve endings. This difference between these types of reflexes can be used to study control of the spindle system (i.e. gamma motoneuron activity). Ib inhibition can be studied by various protocols of double shock conditioning of the H reflex. In conditioned H reflex protocols, Ib inhibition lasts less than 10 ms. Various physiological aspects, such as pre-synaptic inhibition, reciprocal inhibition, facilitation of Ia afferents and Group II afferents can also be studied and quantified in Table 1 Ashworth scale Grade Patient’s status 0 No increase in tone 1 Slight increase in tone with a catch and release or minimal resistance at end of range 1+ Same as 2 but with minimal resistance through range following catch 2 More marked increase in tone but limb easily flexed 3 Considerable increase in tone, through range of motion 4 Affected part rigid Table 2 Tardieu scale Grade Feature 0 No resistance throughout the course of passive movement 1 Slight resistance throughout the course of the passive movement with no clear catch at a precise angle 2 Clear catch at a precise angle, interrupting the passive movement, followed by a release 3 Clonus fatigable (less than 10 s, when maintaining pressure); appears at a precise angle 4 Clonus indefatigable (less than 10 s, when maintaining pressure); appears at a precise angle Childs Nerv Syst (2007) 23:957–970 969 relation to the H reflex to give a more precise electrophys- iological diagnosis and assessment of spasticity. Hyperexcitability of the stretch reflex in spastic patients is characterised by an increase in the Hmax/Mmax ratio because of exaggerated facilitation of the H reflex to voluntary contraction and to the absence of inhibition associated with relaxation. Disynaptic Ib inhibition appears to be depressed in patients with spasticity related to a brain lesion but not in patients with spinal cord lesions. Discordant results accord- ing to the origin of spasticity have also been reported for most of the other electrophysiological recordings. References 1. Banks HH, Green WT (1960) Adductor myotomy and obturator neurotomy for the correction of adductive contracture of the hip in cerebral palsy. J Bone Jt Surg Am Vol 42:111–126 2. Berard C, Sindou M, Berard J, Carrier H (1998) Selective neurotomy of the tibial nerve in the spastic hemiplegic child: an explanation of the recurrence. J Pediatr Orthop B 7:66–70 3. Brunelli G, Brunelli F (1983) Selective microsurgical denervation in spastic paralysis. Ann Chir Main 2:277–280 4. Cahuzac M, Mansant C (1968) Le traitement chirurgical et ses résultats dans le pied spastique de l’enfant. Ann Podol 6:95–105 5. Decq P (2003) Peripheral neurotomies for the treatment of focal spasticity of the limbs. Neurochirurgie 49:293–305 6. Decq P, Filipetti P, Feve A, Saraoui A (1996) Neurotomies périphériques sélectives des collatérales pour les muscles ischio- jambiers dans le traitement de la spasticité en flexion du genou. A propos d’une série de 11 patients. Neurochirurgie 42:275–280 7. Decq P, Filipetti P, Feve A, Djindjian M, Saraoui A, Keravel Y (1997) Peripheral selective neurotomies of the brachial plexus branches for the spastic shoulder. Anatomical study and clinical results in 5 patients. J Neurosurg 86:648–653 8. Decq P, Cuny E, Filipetti P, Feve A, Keravel Y (1998) Peripheral neurotomy in the treatment of spasticity. Indications, techniques and results in the lower limbs. Neurochirurgie 44:175–182 9. Decq P, Cuny E, Filipetti P, Keravel Y (1998) Role of soleus muscle in spastic equinus foot. Lancet 11(352):118 10. Decq P, Filipetti P, Cubillos A, Slavov V, Lefaucheur JP, Nguyen JP (2000) Soleus neurotomy for treatment of the spastic equinus foot. Groupe d’Evaluation et de Traitement de la Spasticite et de la Dystonie. Neurosurgery 47:1154–1160 11. Deltombe T, Gustin T, Laloux P, De Cloedt P, De Wispelaere JF, Hanson P (2001) Selective fasicular neurotomy for spastic equinovarus foot deformity in cerebral palsy children. Acta Orthop Belg 67:1–5 12. Dengler R, Konstanzer A, Hesse S et al (1989) Collateral nerve sprouting and twitch forces of single motor units in conditions with partial denervation in man. Neurosci Lett 97:118–122 13. Feve A, Decq P, Filipetti P, Verroust J, Harf A, Nguyen JP, Keravel Y (1997) Physiological effects of selective tibial neurotomy on lower limb spasticity. J Neurol Neurosurg Psychiatr 63:575–578 14. Filipetti P, Decq P, Feve A, Kolanowski E, Deltombe T (1998) Blocs moteurs périphériques et restauration fonctionnelle. a propos de 202 blocs. Ann Réadapt Méd Phys 41:1–8 15. Garland DE, Thompson R, Waters RL (1981) Musculocutaneous neurotomy for spastic elbow in flexion in non-functional upper extremities in adults. J Bone Jt Surg Am Vol 62:108–112 16. Gros C (1972) La chirurgie de la spasticité. Neurochirurgie 23:316–388 17. Gros C, Frerebeau P, Benezech L et al (1977) Selective radicular neurotomy. In: Simon L (ed) Actualities in functional physical therapy. Masson, Paris, pp 230–235 18. Hodgkinson I, Sindou M (2004) Decision making for treatment of disabling spasticity in children in “Neurosurgical Management of Disabling Spasticity” (Sindou ed.). Oper Tech Neurosurg 7:120–123 19. Hopkins WG, Slack JR (1981) The sequential development of nodal sprouts in mouse muscles in response to nerve degeneration. J Neurocytol 10:537–556 20. Keats S (1957) Combined adductor–gracilis tenotomy and selective obturator nerve resection for the correction of adduction deformity of the hips in children with cerebral palsy. J Bone Jt Surg 39:1087–1090 21. Lorenz F (1887) Über chirurgische behandlung der angeborenen spastischen gliedstarre. Wien Klin Rundsch 21:25–27 22. Mertens P, Sindou M (1991) Selective peripheral neurotomies for the treatment of spasticity. In: Sindou M, Abbott R, Keravel Y (eds) Neurosurgery for spasticity. A multidisciplinary approach. Springer, New York, pp 119–132 23. Mertens P, Sindou M (2001) Surgical management of spasticity. In: Barnes MP, Johnson GR (eds) Clinical management of spasticity. Cambridge Univ. Press, Cambridge, pp 239–265 24. Park DM, Shon SK, Kim YJ (2000) Direct muscle neurotization in rat soleus muscle. J Reconstr Microsurg 16:379–383 25. Purohit AK, Raju BSV, Shiv Kumar K et al (1998) Selective musculocutaneous fasciculotomy for spastic elbow in cerebral palsy: a preliminary study. Acta Neurochir (Wien) 140:473–478 26. Roujeau T, Lefaucheur JP, Slavov V, Gherardi R, Decq P (2003) Long-term course of the H-reflex after selective tibial neurotomy. J Neurol Neurosurg Psychiatr 74:913–917 27. Rotshenker S (1978) Sprouting of intact motor neurons induced by neuronal lesion in the absence of denervated muscle fibers and degenerating axons. Brain Res 155:354–356 28. Russell DJ, Rosenbaum PL, Avery LM, Lane M (ed) (2002) Gross motor function measure (GMFM-66 & GMFM-88). Mac Keith, Cambridge 29. Sindou M, Mertens P (1988) Selective neurotomy of the tibial nerve for the treatment of the spastic foot. Neurosurgery 23: 738– 744 30. Stoffel A (1912) The treatment of spastic contractures. Am J Orthop Surg 10:611–644 31. Tardieu G (1983) Spasticity: an ill-defined condition abstract retrospective neurology in the central and peripheral nervous system. Venice World Federation Neurol, May, pp 24–26 970 Childs Nerv Syst (2007) 23:957–970 Selective peripheral neurotomy (SPN) for spasticity in childhood Abstract Abstract Abstract Abstract Introduction Principles Rationale Preoperative assessment Therapeutic objective Clinical examination Motor blocks Technical basis of neurotomy Anaesthesia Mapping Section Operative techniques Surgery for lower limb Obturator neurotomy for the hip Hamstring neurotomy for the knee Tibial neurotomy for the foot Anterior tibial neurotomy for extensor hallucis Femoral neurotomy for the quadriceps Surgery for the upper limb Pectoralis major neurotomy for the shoulder Teres major neurotomy for the shoulder Musculo-cutaneous neurotomy for the elbow Median neurotomy for the wrist and fingers Ulnar neurotomy for wrist and fingers Complications, side effects and causes of recurrence Indications General indications Predictability of outcome Conclusions Appendix&newnbsp;1: The concept of focal spasticity Appendix&newnbsp;2: Preoperative assessment Appendix&newnbsp;3: Electrophysiology References /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /Warning /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 150 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 600 /MonoImageMinResolutionPolicy /Warning /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 600 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description > >> setdistillerparams > setpagedevice