Neurosurgical lesioning-procedures for spasticity and focal dystonia

35.2.1

Peripheral neurotomies

Peripheral neurotomy (PN) is indicated when spasticity is focal, that is, localized to a few muscular groups supplied by one single or a few peripheral nerves, only after repeated botulinum toxin injections are revealed insufficient . PN consists in partial sectioning of the motor fascicles corresponding to the muscles harboring excessive spasticity; the procedure acts by interrupting the segmental reflex arch at both its afferent and efferent pathways. PN must not involve sensory fibers, as their damaging would result in paresthesias and secondary neuropathic pain. Empirically it is agreed upon that 60% of all motor fascicles to the targeted muscles have to be sectioned to be effective. Surgery aims not only to reduce spasticity but also to improve function by reequilibrating the tonic balance between agonist and antagonist muscles as shown in Fig. 35.2 .

Tibial neurotomy for the spastic dystonic foot and obturator neurotomy for spastic flexion–adduction of the hip are the most common neurotomies. PN can also be indicated for spasticity/focal dystonia in the upper limb: musculocutaneous neurotomy for elbow in flexion, median (and ulnar) neurotomy(ies) for flexion of the wrist and fingers to improve at least cosmetics and prehension if hand retains residual motor function.

Before deciding PN, local nerve blocks with a (2–3 hours) long-lasting anesthetic (bupivacaine among them) can be useful to evaluate the possible hidden motor power of the antagonist muscles and to determine whether the eventual articular limitations result from spasticity or musculotendinous contractures/articular ankyloses. Also, local blocks, as well as botulinum toxin injections (being used as a “prolonged” test of several months), by mimicking the effect of corresponding neurotomy(ies), allow the patients and the medical team to estimate the benefit that would follow the surgery.

Intraoperative mapping is fundamental to achieve effective and selective PN. General anesthesia must be induced without long-lasting myorelaxants so that the muscular responses to electrical stimulation can be detected. Nitrous oxide and propofol should be avoided because they modify reflex excitability. Use of the operating microscope for precise dissection is recommended ( Fig. 35.3 ) . Individual mapping of branches/fascicles is required, because of the frequent variations in the emergence and distribution of the nerve branches. Bipolar stimulation is preferred to monopolar stimulation to avoid spread of the electric current. Stimulation (such as the one using the Innopsys Nimbus I-Care multifunctional neurostimulator, manufactured by Innopsys, Parc d’Activités Activestre, 31390, Carbonne, France, www.innopsys.com ) is performed with a frequency of 2 Hz and at a (just above the threshold) intensity (currently 200 μA) to avoid electric diffusion and, therefore, incorrect interpretation. When fascicles are tightly apposed, a triple stimulation probe composed of an anode between two cathodes is the more accurate way of stimulating because of less current diffusion. The responses to stimulation can be visualized in the form of clinically observable muscular twitches in the corresponding limb. The surgeon may also have recourse to electromyogram (EMG) recordings. Extreme care should be taken to recognize false-positive responses by diffusion of current to neighboring fascicles and false-negative responses due to short circuits of current caused by serosities in between the contacts of the stimulation probe.

Once all motor branches/fascicles have been identified, those considered to be targeted are marked separately with different-colored threads. According to preoperative program, variable proportions (60%–80% depending on the degree of preoperative spasticity) of the selected motor branches/fascicles are cut. The effect of each sectioning is then evaluated by comparing the intensity of the muscular response to electric stimulation, proximally and then distally to the section. If the response after proximal stimulation is still intense, further sectioning is performed.

35.2.2

Dorsal rhizotomies

Based on Sherrington’s experiments, Foerster performed the first lumbo-sacral dorsal rhizotomies (DRhs) in cerebral palsy (CP) patients. As a general rule he recommended “to resect the dorsal roots from L2 to S2 with the exception of the fourth lumbar root since this root generally guarantees the extensor reflex of the knee, so very necessary for standing and walking” . To preserve further the innervation of the muscles responsible for “useful tone,” the quadriceps and abdominal and gluteal muscles in particular, Gros introduced in 1976 topographic selection of the roots by their electrical stimulation, technique termed sectorial posterior rhizotomy . Apart from effects on the lower limbs, Gros observed in this CP patients decrease in spasticity in the upper limbs and improvement in speech and swallowing, so-called distant effects. Also in 1976 Fasano et al. developed the functional posterior rhizotomy, a method based on stimulation of the DR/dorsal rootlets with corresponding EMG recordings . Exaggeration of the duration or of the extent of the muscular responses indicates the particular roots that must be surgically sectioned.

At present, the most used techniques are based on the Fasano’s principle of determining the DRs/rootlets to be cut from the intensity of the muscular responses to their stimulation. Roots/rootlets can be accessed through two different types of approaches: a one-piece laminotomy performed from L1 to S1, as developed by Peacock and Arens and Abbott et al. , or a limited laminotomy of one/three Th11–L1 vertebrae in the front end of the conus medullaris, as popularized by Park et al. .

To reduce invasiveness of the intraspinal approach via laminotomies, and at the same time access each of the lumbosacral roots—individually—at their corresponding dural sheath, we developed what we termed “keyhole interlaminar DRh” ( Fig. 35.4 ) . With this technique, each enlarged interlaminar space enables to access intradurally, two roots on one side (one upward and one downward). Steps are the following ( Fig. 35.5 ) . The L1–S2 lumbosacral spine is exposed posteriorly through a midline skin incision from Th12 to S1. The chosen interlaminar spaces (two, or more often three, according to patient’s clinical presentation and therefore roots selected to be targeted) are extended in height by resecting the lower two-thirds of the upper lamina and the upper three-fourths of lower lamina. The spinous processes and their interspinous ligaments are respected. Through the obtained fenestration(s), flavum ligament is resected, and dura and arachnoid are incised on the midline for a height of approximately 2 cm. The L2 and L3 roots can be reached via a L1–L2 opening, L4 and L5 via a L3–L4 opening and S1 and S2 via a L5–S1 opening.

Then the microscope is installed for a trajectory oblique at approximately 45 degrees, so that following the principles of keyhole surgery, the surgeon’s view passes underneath the (respected) interspinous ligament. The goal is to access—intradurally—the contralateral roots, with their ventral and dorsal components, respectively, at exit to/entry from the corresponding dural sheath. At each level, muscle responses to stimulation are checked using a bipolar electrode to avoid spreading of current. Ventral root (VR) is stimulated first, to identify its innervation territory and thereby confirm topographical level; this phase is “anatomical mapping.” Stimulation is at 2 Hz with an intensity of 200 μΑ (to 500 μA), that is, slightly above threshold. To be noted, a muscle response by DR stimulation would require three to five times higher intensity. Then the DR undergoes “physiological testing” to evaluate excitability of its contained circuitry. Stimulation is a train of 1 second at frequency of 50 Hz and with an intensity of 1 mA, not to have excessive diffusion of current. Excitability is considered “exaggerated,” according to Fasano’s classification , when the muscular responses are sustained and/or spreading out “normal” myotomes. This testing is to confirm or modify the percentage of the dorsal rootlets, previously specified in the preoperative chart, to be cut. The amount of cut generally ranges between one-third and four-fifths of a root’s constituting rootlets. Neurophysiological mapping and testing ( Fig. 35.5 ) are followed by the microsurgical sectioning of each monitored root. Noteworthy, same steps are performed at each targeted radicular level, on one side first, then on the other side.

To be effective there is an empirical consensus that an overall amount of 60% of all the targeted DRs should be cut, with a different proportion according to the level and function of the roots involved and, of course, their implication in the “conveying” of the harmful components of the spasticity .

An overall amount of 60% of cutting is compatible with the absence of lasting sensory deficits, as shown by a long-term study of somatosensory evoked potentials (SEPs) after lumbosacral DRh .