Selective Dorsal Rhizotomy and the Spasticity Ladder

Evidence status. SDR reliably reduces lower-extremity tone in selected children with cerebral palsy. Functional gains and long-term comparative advantage are more variable; evidence is largely observational and selection-dependent.

1.Define goals and the motor phenotype

Goals may be active (walking efficiency, speed, balance, orthotic tolerance) or passive (comfort, sleep, hygiene, dressing, seating, and ease of care). Analyze gait, strength, selective control, range, torsion, contracture, dystonia, cognition, communication, motivation, family capacity, and access to intensive rehabilitation.

Use standardized measures such as GMFCS, the GMFM (record which version: GMFM-66 and GMFM-88 are not interchangeable, and the older randomized literature and the modern meta-analyses report different ones), modified Ashworth or Tardieu measures, instrumented gait when it will change a decision, and patient/family goal attainment. Tone scores alone do not establish benefit.

2.Escalate by reversibility and distribution

3.Mechanism and techniques

SDR divides selected lumbosacral dorsal rootlets to reduce afferent drive into hyperexcitable spinal reflex circuits. Multilevel cauda-equina and limited conus-level exposures differ in root identification, bony exposure, and monitoring. The conus level varies and must be localized from patient imaging, often supplemented by intraoperative ultrasound; a fixed vertebral level is not a safe substitute. The proportion divided varies by technique, level, phenotype, and institutional protocol. More extensive section is not automatically better: excessive reduction can expose weakness and impair function. No universal rootlet percentage predicts long-term benefit.

Intraoperative stimulation helps distinguish dorsal from ventral structures and characterize reflex spread, but the stimulation train, grading system, and selection thresholds vary by center. Anatomy, reproducible physiological responses, and the planned distribution of tone reduction must agree before division. Sacral and sphincter monitoring informs preservation of continence-related pathways, but does not guarantee bladder or bowel safety. Do not identify sacral levels or continence pathways from a fixed distance at the conus alone; individual anatomy and overlapping sacral innervation require specialist verification.

The irreversibility belongs in every discussion. The immediate reduction in tone is not the end of the operation but the beginning of a rehabilitation problem: previously concealed weakness may become apparent. Discuss anticipated temporary weakness while recognizing that severe or persistent loss of function is an adverse outcome requiring assessment. Two consequences deserve naming. In the ambulatory child, removing extensor tone from a quadriceps that cannot substitute for it can produce or deepen crouch, which is why the postoperative strengthening program is part of the operation and not an afterthought. In the child who has been weight-bearing through tone to make a standing transfer possible, removing that tone can cost the family the transfer. That is precisely why comfort-focused rhizotomy needs its own consent conversation and its own endpoints.

4.Classic ambulatory candidate

The clearest traditional candidate has bilateral lower-extremity spasticity from cerebral palsy, usually GMFCS II or III, adequate antigravity strength and selective control, minimal dystonia, manageable contracture, preserved family participation, and access to prolonged postoperative therapy. Age windows vary by program, but most cluster between three and ten years, and the published English commissioning criteria (three to nine years inclusive, GMFCS II or III, dynamic lower-limb spasticity affecting function, no dystonia, an MRI showing typical cerebral palsy change without basal ganglia or cerebellar damage, a Reimers migration index below forty percent, no significant scoliosis, weakness no worse than moderate with antigravity postures preserved, no progressive disease, and a committed rehabilitation pathway) are a reasonable template for what the outcome literature is actually conditioned on. The published criteria have themselves been systematically reviewed; read that review rather than any single program protocol. These criteria describe a classic ambulatory pathway, not every cohort on this page. Match each outcome estimate to the study’s age, GMFCS level, phenotype, technique, and rehabilitation exposure.

Brain MRI pattern, hip surveillance, spine and limb alignment, gait analysis, prior toxin response, and orthopedic sequencing help forecast the pathway. Significant dystonia, profound weakness, absent selective control, or unrealistic gait goals reduce the case for SDR. Reducing the case for rhizotomy is not the same as having nothing to offer. Where the dystonic component dominates, the conversation moves to intrathecal baclofen, which treats mixed tone, or to pallidal stimulation. Before quoting any number, read the dyskinetic cerebral palsy section of the dystonia page in this library, because the motor gain there is much smaller than in isolated dystonia and much of the residual disability is fixed rather than dystonic. Combined anterior and posterior rhizotomy has been described for mixed spastic-dystonic lower limbs; the reported series are small and the technique is not standard.

5.Nonambulatory and comfort-focused SDR

Some centers use SDR in GMFCS IV-V patients to improve comfort, hygiene, positioning, and care rather than gait. This is a different indication with different outcomes. Compare it explicitly with intrathecal baclofen, orthopedic care, and ongoing focal treatment, and be honest that the head-to-head literature comparing rhizotomy with pump implantation in children amounts to non-randomized institutional series rather than trials. Avoid importing ambulatory GMFM claims into a comfort-focused consent.

6.What the evidence supports

Long-term studies consistently show reduced lower-extremity tone. The functional claim is weaker than the tone claim, and it matters which kind of number is being quoted. Controlled evidence amounts to three randomized trials from the 1990s, pooled in a meta-analysis that put rhizotomy with physiotherapy roughly four GMFM percentage points ahead of physiotherapy alone. The contemporary meta-analyses are single-arm pools of before-and-after data: the most recent, sixteen studies and 756 children, reports a mean GMFM-66 gain near 3.3 points at earliest follow-up and 3.7 at latest, with heterogeneity above seventy percent and every included study at moderate risk of bias. A before-and-after pool in a growing child receiving months of intensive therapy cannot separate the rhizotomy from growth and rehabilitation, and the honest reading is that it supports an association with functional improvement and tone reduction without isolating the procedure’s causal contribution. Comparative proof that SDR is superior to equally intensive multidisciplinary care over decades remains absent. Many patients later need orthopedic surgery or additional tone management.

Report individual trajectories, GMFCS-stratified outcomes, energy and participation, pain, hip and spine surgery, further tone procedures, and adverse events. A mean score does not capture a child who walks farther but develops foot deformity, or one whose caregiving improves without GMFM change.

7.Risks and sequencing

Separate the complications of the operation from the natural history it does not alter. Sensory change, dysesthesia, neuropathic pain, transient urinary retention, CSF leak, wound infection, and unmasked weakness belong to the operation; historical series reported transient dysesthesia in a substantial minority of children, with urinary retention and permanent sensory loss less common; those figures predate contemporary exposures and monitoring but frame the conversation, and a center should quote its own audited rates. Hip displacement, by contrast, largely reflects the underlying cerebral palsy phenotype and should be neither automatically blamed on nor excused from the surgery. In long-term follow-up of ninety children at a mean of eight and a half years, migration percentage exceeded thirty percent in eight percent at GMFCS II, forty-six percent at III, seventy-two percent at IV, and seventy-eight percent at V, matching population natural history by level, and the authors interpreted these observational findings as consistent with no protective or harmful effect. A retrospective comparison with population rates cannot establish absence of an effect in every patient. Rhizotomy is not hip surgery and does not substitute for hip surveillance, which continues on the GMFCS-based schedule afterwards exactly as it would have without it.

Spinal deformity has two separate stories in this literature and they should not be merged. Five-level lumbosacral laminectomy was followed by isthmic spondylolysis or grade I spondylolisthesis in a fifth of the ninety-nine children re-imaged in one early Cape Town cohort (asymptomatic, without further slip during that follow-up), and findings of this kind pushed programs toward narrower exposures; the argument that a single-level or osteoplastic exposure disrupts less is anatomically sound but has not been established against multilevel laminectomy in a controlled comparison. Set against that, radiographic follow-up of a hundred and seven children at a mean of nearly thirteen years found scoliosis (a curve above ten degrees) in just over half, concentrated at the nonambulatory levels (about a third at GMFCS II and under a fifth at III, against nearly three quarters at IV and more than nine in ten at V), with spondylolisthesis on sixteen percent of the lateral films available; those authors read the rates as the natural history of cerebral palsy rather than a consequence of rhizotomy. Surveil the spine for life; do not promise a family that a smaller exposure removes the risk.

Do not combine SDR and major orthopedic correction merely for convenience unless the rehabilitation logic is explicit. Tone reduction can reveal alignment and strength needs; staged reassessment often clarifies the definitive orthopedic plan.

8.Rehabilitation is part of the operation

Define prehabilitation, inpatient expectations, frequency and duration of therapy, home program, orthoses, strengthening, gait retraining, school needs, and travel burden before surgery. If that pathway is unavailable, the operation being consented is not the operation represented in published high-resource cohorts.