Selective Dorsal Rhizotomy (SDR) for Spastic Cerebral Palsy

Indication. Selective dorsal rhizotomy is indicated for spastic diplegia cerebral palsy (bilateral lower-limb spasticity) in ambulatory children (GMFCS level II-III), typically aged 3-8 years, in whom spasticity predominates over weakness (Modified Ashworth Scale 2 or more) and conservative management has failed. The procedure interrupts the sensory limb of the exaggerated stretch reflex by dividing a proportion of the lumbosacral dorsal rootlets, permanently reducing spasticity while preserving useful motor control and sensation. The right child. Beyond the diagnosis, the candidate must have: - Predominant spasticity over weakness, with good selective motor control and dynamic spasticity that improves with nerve blocks or sleep.

• Adequate trunk control and upper-limb function, and a motivated child and family able to commit to intensive post-operative rehabilitation.
• No fixed contractures requiring extensive orthopaedic surgery first (small, flexible contractures may be addressed later). Failed conservative management must be documented — intensive physiotherapy (at least 12 months), appropriate orthoses (AFOs or KAFOs), multiple cycles of botulinum toxin, and oral anti-spasmodics (baclofen, diazepam).

Pre-operative assessment. Confirm the movement disorder is spasticity, not dystonia. Document the GMFCS level and selective motor control, grade spasticity (Modified Ashworth) and strength (MRC), and assess range of motion. Perform 3D gait analysis to quantify the spasticity distribution and overactive muscle groups — it directly guides the level-by-level intra-operative plan. Image the spine with MRI to exclude a tethered cord, syrinx or Chiari malformation, and obtain hip radiographs to assess the migration percentage. Consent specifically for transient sensory change or dysesthesia (5-10 percent), CSF leak or pseudomeningocele (5-8 percent), post-laminectomy spinal deformity (10-15 percent kyphosis), a small risk of new motor weakness from ventral root injury (less than 2 percent), temporary bladder dysfunction, and the need for prolonged intensive physiotherapy and possible later orthopaedic surgery. Set realistic expectations: SDR reduces spasticity durably but does not cure cerebral palsy, does not prevent fixed contractures, and the added gross-motor benefit over high-quality physiotherapy alone is modest. Setup. Prone on a radiolucent Jackson table with the abdomen hanging free and the hips flexed 30-45 degrees (table break) to open the interspinous spaces and reduce epidural venous engorgement. Comprehensive neuromonitoring — SSEPs, MEPs and multi-channel EMG — is essential, and baselines are recorded before draping.

The goal is to expose the lumbosacral thecal sac through a posterior midline approach, open the dura in strict midline, identify each dorsal root from L2 to S2 and separate it into rootlets, then use intra-operative EMG to select and divide the abnormal (spasticity-generating) rootlets while preserving motor function and sensation. The exposure — positioning, incision, muscle elevation, the bone-sparing single-level L1 laminectomy and the dural opening — is laid out in full as the first steps below, because safe access to the cauda equina is the foundation of the whole operation.

View larger

Rehabilitation — the operation creates a window of opportunity by removing spasticity; intensive therapy is what converts that into functional gain. | Phase | Timing | Focus | |-------|--------|-------| | 1 | 0-2 days | ICU or HDU, neuro observations every 2 hours, flat bed rest 24-48 hours, Foley catheter, multimodal analgesia | | 2 | Day 2-5 | Sit, stand and begin gentle range of motion; early mobilisation capitalises on the tone reduction; AFO review | | 3 | Week 1 to month 6 | Intensive physiotherapy 3-5 sessions per week: quadriceps and dorsiflexor strengthening, balance and proprioception, gait re-education | | 4 | Month 6 to year 2 | Repeat 3D gait analysis at 3 and 6 months; orthopaedic review for residual fixed contractures; annual spine radiographs | Bladder retention is common and usually temporary; a persistent deficit beyond 48 hours warrants evaluation for a cauda equina problem. Most children return to their pre-operative mobility level within weeks, with durable tone reduction and gradual functional gains over the first 1-3 years. Complications

Rationale and history. Spasticity is velocity-dependent increased tone driven by an exaggerated stretch reflex. SDR reduces it by dividing a proportion of the lumbosacral dorsal (sensory) rootlets, interrupting the sensory afferent limb of that reflex at the spinal cord while sparing the motor (ventral) roots. The technique was developed by Fasano and refined by Peacock and Arens (Cape Town, 1982), then evolved into the modern bone-sparing single-level conus laminectomy described by Park and Johnston. The spinal cord and cauda equina. The conus medullaris terminates at L1-L2 in adults but may extend to L3 in young children (and to L3-L4 in neonates) — its level must be confirmed before laminectomy so the bone window sits at or below the conus. Below the conus, the L1-S5 nerve roots descend within the thecal sac as the cauda equina, becoming progressively more oblique (the L2 root runs at about 30 degrees, the S1 root at about 60 degrees to the horizontal). Each dorsal root divides into 3-7 rootlets (average 4-5), which are the units tested and divided. Blood supply. A single midline anterior spinal artery supplies the anterior two-thirds of the cord (including the motor tracts) and is fed by 6-10 radicular arteries along its length, including the dominant artery of Adamkiewicz (entering T9-L2, usually on the left) which supplies the lumbar enlargement — its injury causes anterior cord syndrome. Paired posterior spinal arteries supply the posterior third. Radicular arteries accompany the roots through the foramina and can bleed during rootlet division, usually controlled with bipolar cautery.

Anatomy at risk

Key evidence. The highest-level evidence is the individual-patient meta-analysis of the three published RCTs (McLaughlin, 2002, PMID 11811645): SDR plus physiotherapy reduced spasticity significantly more than physiotherapy alone, with a small but real additive gross-motor benefit (about 4 GMFM points) and a dose-response between the percentage of root tissue transected and functional gain. Wright (1998, PMID 9593495) showed that spasticity reduction translates into measurable gait improvement at one year. Selection remains the single most important determinant of outcome and is non-standardised worldwide (Grunt, 2014, PMID 24106928). Long-term, the tone reduction is durable for decades but gross motor function peaks around 3 years then gradually declines, and SDR does not prevent contractures (Tedroff, 2015, PMID 25523506). Spinal deformity after modern single-level SDR is driven mainly by baseline patient factors rather than the laminectomy itself (Ravindra, 2017, PMID 28885083).