Before the needle: one question that sorts everything
By now you have the gentle end of the toolkit firmly in hand — daily stretching, careful positioning, splints and serial casting, and the oral antispasticity pills. You also carry the most important lesson of this whole rung: the aim of tone management is not to abolish stiffness but to weigh, joint by joint, where reducing it helps and where the very same tone is quietly doing useful work. The tools in this guide are the sharper, more powerful ones you reach for when the gentle measures are not enough — and precisely because they are powerful, that weighing matters more, not less. A bigger lever can pry open more, and it can also break more.
One question sorts the whole toolkit before you choose anything: is this spasticity focal or generalized? Focal means the problem lives in one or two specific muscles — a thumb curled tight into the palm, a calf that points the foot down so the toe scuffs the floor at every step. Generalized means stiffness is spread across many muscles at once — both legs after a spinal cord injury, much of the body after multiple sclerosis. The match is almost a rule. A focal problem wants a focal tool that touches that muscle and leaves the rest of the body untouched. A generalized problem cannot be solved one muscle at a time, so it wants something that reaches everywhere — but that reach is exactly what costs you in side effects.
Botulinum toxin: switching off one muscle at the muscle
For the focal problem, the elegant answer is to switch off just that one muscle, right at the muscle, by injecting a tiny dose of botulinum toxin into it. The same toxin that in food poisoning paralyses muscles all over the body is here turned into a precision instrument: a minute amount, placed exactly where you want weakness and nowhere else. The technical name for what it does is chemodenervation — a chemical that temporarily disconnects the muscle from its nerve. Think of it as quietly unplugging one overactive appliance instead of throwing the whole building's breaker, which is what an oral pill does.
The mechanism is worth holding precisely, because it explains both the gift and the limits. At the neuromuscular junction — the tiny gap where a nerve hands the muscle its "contract now" message by releasing a chemical messenger called acetylcholine — botulinum toxin blocks that release. The message simply never arrives, so the injected muscle goes weak and quiet. The effect stays mostly where it is placed (it is local), and it is temporary: over roughly three to four months the nerve endings sprout new connections, the muscle wakes back up, and the injection must be repeated. That fading is not a flaw to lament; it is a window to exploit. While the overactive muscle is hushed, the limb is stretched, cast, and retrained — so the months of quiet are spent locking in gains the spasticity had been blocking.
Be honest about what it cannot do, because the misconceptions are common. It works only on dynamic, reflex-driven overactivity — a live, switchable signal; it does nothing for a fixed contracture. The dose into each muscle and the total dose are both capped, so you cannot quiet many large muscles at once with it. It must be repeated for as long as the spasticity lasts. And it is most effective paired with therapy: injecting without the stretching and retraining that follow wastes most of the benefit, like loosening a stubborn bolt and then walking away before turning it. Finding the right muscle — often with ultrasound or electrical-stimulation guidance to confirm the needle is truly in the troublemaker — is a real part of the skill.
Phenol and alcohol: the older, blunter tool for big muscles
Botulinum toxin is precise and gentle, but it is costly, capped by dose, and wears off in a season. For a large, powerful muscle group driving severe stiffness — the big muscles that pull the hip or the knee — there is an older, blunter, far cheaper tool: drop phenol or alcohol directly onto the nerve that feeds the muscle, chemically damaging it so it stops carrying the overactive signal. This is neurolysis — literally "loosening the nerve" — and it long predates botulinum toxin. Where the toxin blocks the message at the muscle, neurolysis attacks the wire itself: dropped onto a motor nerve, phenol destroys part of its insulation and fibres, so fewer signals get through and the muscle relaxes.
The trade-offs run in mirror image to the toxin, which is why the two are partners more than rivals. Neurolysis comes on within minutes rather than days, can last far longer — many months, sometimes a year or more — costs a fraction as much, and carries no total-dose ceiling, so it suits the big muscles a toxin budget cannot afford. The price is precision and a particular risk: many nerves carry sensation as well as motor signals, and phenol damages whatever the nerve carries. Hit a mixed nerve and you can leave a lasting burning dysaesthesia — a painful, abnormal sensation. The technique is also more uncomfortable and more operator-dependent, demanding careful electrical stimulation to find the right nerve before the chemical is placed.
The intrathecal baclofen pump: delivery to where it is needed
Now switch from the focal column to the generalized one. When stiffness is spread across both legs and the trunk, a focal injection cannot keep up — there are simply too many muscles. The natural reach-everywhere answer is an oral pill like baclofen, but you have already met its catch among the oral antispasticity agents: baclofen calms tone by acting in the spinal cord, yet swallowed as a tablet it must first flood the whole body and brain to get a little of itself down to the cord. That is why oral baclofen so often fogs and drowses a person before it loosens the legs — most of the drug is reaching the brain, where you did not want it.
The intrathecal baclofen pump solves this by changing the delivery, not the drug: a thin tube drips baclofen directly into the fluid bathing the spinal cord, so a tiny fraction of the oral dose does the same job with far less reaching the brain. The system is fully implanted. A hockey-puck-sized pump holding a reservoir of the drug sits under the skin of the abdomen; a fine catheter runs from it around to the spine and into the intrathecal space — the fluid-filled sheath around the cord. The pump is programmed to release a steady, adjustable micro-dose, and the reservoir is refilled with a needle through the skin every few months. Because the dose at the cord is high while the dose to the brain stays low, it can powerfully quiet severe, widespread lower-body spasticity while leaving the person far more alert than oral medication ever could.
The pump is reserved for severe, generalized lower-body spasticity that oral drugs cannot control without intolerable sedation — often after spinal cord injury, cerebral palsy, or multiple sclerosis — and candidates usually get a trial test dose first to confirm it truly helps before anything is implanted. The honest cautions are serious. It is surgery, with the usual risks plus device problems: a catheter can kink, a pump can fail, an implant can become infected. The gravest hazard is sudden withdrawal: if delivery stops unexpectedly, baclofen levels crash and can cause a life-threatening rebound of spasticity, high fever, and instability. A pump is not a one-time procedure but a lifelong commitment to refills and monitoring.
Surgery: the permanent answers at the bottom of the ladder
Everything so far has worked on a moving target — a nerve signal turned down for a while, then re-treated. But some problems have become permanent, and some severe spasticity outlasts every gentler tool. Then surgery offers a more definitive answer, sitting at the bottom rung of the ladder for sober reasons: it is irreversible, it carries operative risk, and it demands months of rehabilitation afterward to convert a corrected anatomy back into useful function. You reach it only after the milder options have been tried and the goal clearly justifies an operation.
Surgery comes in two broad families that answer two different problems. Orthopaedic surgery works on the muscles, tendons, and bones themselves, and it is the answer for a true fixed contracture that no drug can touch: a permanently shortened tendon can be lengthened so the joint regains range, a tight muscle released, an overactive muscle transferred to pull in a more useful direction, a fixed deformed bone cut and realigned. A classic example is lengthening a chronically short Achilles tendon so a child with cerebral palsy can finally get the heel down to the floor. Neurosurgery instead works on the nerves: in selective dorsal rhizotomy, the surgeon cuts a carefully chosen fraction of the sensory nerve rootlets entering the lower spinal cord — the very rootlets that feed the overactive reflex — permanently reducing spasticity in the legs.
Selective dorsal rhizotomy is best evidenced in younger children with spastic diplegia who already have reasonable underlying strength and few fixed contractures, and the selection is the whole game — it is not for dyskinetic or ataxic patterns, and it is weighed carefully against its irreversibility. Here the recurring lesson of this whole field lands hardest. Removing spasticity unmasks whatever strength lies beneath; if useful supporting tone was being borrowed to stand, taking it away can leave a child looser but unable to bear weight. Surgery changes the mechanics, but it does not restore the lost strength or control behind the upper motor neuron syndrome — and intensive physiotherapy over the following months is what turns a corrected limb into a functioning one. The honest summary: surgery durably reduces stiffness, but it cannot create strength or coordination the person never had.
Matching the tool to the problem
Step back and the whole toolkit lines up along two axes you already know: focal versus generalized, and reversible nerve signal versus fixed tissue. A focal, live problem in a delicate place wants botulinum toxin; in a big powerful muscle, phenol neurolysis. A generalized, live problem of the lower body wants the intrathecal baclofen pump (or oral drugs, if mild enough to tolerate). And a fixed contracture — anything that will not move even when the person is asleep — bypasses all of these and goes to mechanical lengthening: casting first, surgery last. These are not rival therapies competing for one patient; they are different keys for different locks, and a single person may carry several locks at once.
MATCHING THE TOOL TO THE PROBLEM PROBLEM TOOL ----------------------------- ---------------------------------- focal, delicate (e.g. hand) botulinum toxin (at the muscle) focal, big muscle (hip/knee) phenol / alcohol neurolysis (nerve) generalized lower body, mild oral antispasticity pills generalized lower body, severe intrathecal baclofen pump (cord) fixed contracture casting -> orthopaedic surgery severe leg spasticity, child selective dorsal rhizotomy axis 1: focal <-> generalized axis 2: live nerve signal <-> fixed tissue always first: what is the GOAL, and is the tone helping or hurting?
Two honesties close the rung. First, no tool here repairs the lesion or restores the lost strength behind the upper motor neuron syndrome — every one of them quiets overactivity, and quieting overactivity adds not one gram of power. That is why each is most useful glued to therapy: the injection or the pump buys a calmer window, and the stretching, casting, and retraining are what convert it into function. Second, the guiding question never changes from the start of this rung to its end. Before reaching for any of these sharper tools, ask what reducing this tone would actually buy the person and what it might cost — because the most powerful instruments are exactly the ones most able to make a weak, borrowing limb worse. The skill is not in wielding the tool. It is in knowing, this time, for this joint, whether to pick it up at all.