Functional Neurosurgery · Trainee Resources

Intrathecal Drug Delivery

Targeted Analgesia for Pain and Intrathecal Baclofen for Spasticity

One implanted system, two distinct therapies: delivering drug straight to its spinal site of action lets a small dose do the work of a far larger systemic one. A practical guide to candidate selection, the trial, the implant, perioperative care, and the complication that must never be missed.

Timoteo Almeida, MD, PhD | Department of Neurological Surgery

Orientation

Intrathecal drug delivery rests on one elegant principle: putting a drug directly into the cerebrospinal fluid, at the spinal segment where it acts, achieves with a tiny dose what would otherwise require a large and toxic systemic one. Intrathecal morphine is roughly three-hundred-fold more potent than oral morphine; intrathecal baclofen reaches the cord at concentrations oral dosing could never safely approach. The payoff is the same in both of the therapy's two worlds — better target effect, far less systemic drug, fewer systemic side effects.

Those two worlds are worth keeping mentally separate even though they share the same pump and catheter. The first is analgesia: refractory cancer pain and selected chronic non-cancer pain, where intrathecal opioid or ziconotide replaces escalating systemic opioids. The second is spasticity: intrathecal baclofen for severe spasticity of cerebral, spinal, or demyelinating origin. The hardware is identical and the operation is nearly the same; the indications, the drugs, the trial, and the emergencies are not. This reading treats them as two parts of one technique, and closes on the single complication — abrupt baclofen withdrawal — that every surgeon must recognize on sight.

Part I

The Shared System and Operation

1.What Is Implanted

The system has two components: a programmable pump — a battery-powered reservoir implanted in a subcutaneous (or subfascial) pocket in the lower abdominal wall — and an intrathecal catheter tunneled subcutaneously from the pump around the flank to a spinal entry point, where it is passed into the lumbar thecal sac and threaded so that its tip sits at the spinal level appropriate to the target. The reservoir is refilled percutaneously through a port at intervals of weeks to months, and the infusion rate is set and adjusted by external telemetry. Catheter-tip position matters: for pain, the tip is placed near the spinal segments serving the painful region; for spasticity, it is advanced higher (commonly mid-to-low thoracic, or higher for predominantly upper-limb spasticity) so baclofen reaches the relevant cord segments.

2.The Trial

Outside of the advanced-cancer setting (where time is short and clinicians may proceed directly to implant), a trial is standard: a test dose or a short externalized infusion confirms that the drug actually helps this patient before a permanent pump is placed.

Pain trial. A single intrathecal bolus, or a short epidural/intrathecal infusion via a temporary catheter, is used to demonstrate meaningful analgesia (conventionally a clear reduction in pain, often taken as ≥ 50%) without intolerable side effects. Representative consensus (Polyanalgesic Consensus Conference) starting points: a morphine bolus trial in the range of 0.2–1.0 mg (more conservatively 0.1–0.2 mg in opioid-naive patients), and for ziconotide a single low-dose bolus on the order of 1–2.5 µg, titrated cautiously given its narrow therapeutic window. These are starting figures only — confirm against current PACC guidance and institutional protocol, and judge success by analgesia and tolerability rather than a fixed dose.
Baclofen trial. A test dose is given by lumbar puncture, the patient positioned lateral decubitus (or sitting), with the drug injected slowly over about a minute after confirming CSF backflow through a small-gauge needle. Start around 50 µg; if there is no adequate response it is repeated on subsequent days at 75 then 100 µg. Assess the response at the drug's peak effect roughly 4 hours after injection, looking for a clear, dose-related fall in tone — conventionally a drop of at least one point on the modified Ashworth scale across the affected muscle groups — which confirms candidacy and is only pursued if a permanent pump is itself a viable option for that patient. The trial also helps separate dynamic spasticity (which baclofen relaxes) from fixed joint contracture or arthrosis after long-standing spasticity (which it cannot release): a limb that fails to gain range after an adequate test dose points to an established contracture — an orthopedic, surgical, or rehabilitation problem, not a pump candidate — which should be ruled out before implant.

A counterintuitive selection caveat Reducing tone is not always desirable. Some patients use their spasticity functionally — extensor tone that lets them stand, transfer, or weight-bear — and abolishing it can leave them worse off. Before trialing, ask whether the spasticity is purely harmful or partly load-bearing; ITB is for spasticity that interferes with care, comfort, or function, not for tone the patient depends on. This is exactly what the test dose helps reveal.

Why the trial is the heart of patient selection The trial converts a judgment call into an observation: you do not implant a pump on the hope of benefit, you implant it after you have watched the drug work in that specific patient. A failed or equivocal trial is a gift — it spares the patient a pocket, a catheter, and a lifetime of refills for a therapy that would not have helped.

3.The Implant Operation, Step by Step

Permanent implantation is performed under general anesthesia (monitored anesthesia care is an option in selected pain patients), with the patient in the lateral decubitus position to give simultaneous access to the spinal midline and the abdominal pocket site, prepped and draped as a single sterile field (an adhesive iodophor drape helps maintain it). Meticulous sterility and meticulous hemostasis — at both the spinal and abdominal wounds — are paramount: an infected pump system almost always has to be explanted, and hematoma/seroma drives wound complications. Fluoroscopy is used throughout, so the table must allow clean AP and lateral images (a radiolucent table, or care that the table support is clear of the field).

Spinal access (paramedian, fluoroscopy-guided). Unlike a standard midline LP, the spinal needle is introduced through a shallow paramedian oblique trajectory — skin entry roughly over the L4 pedicle, advanced toward the midline at about a 30° angle to enter the thecal sac around L2–L3. This oblique path lets the catheter turn cranially without kinking and gives a solid deep-fascial fixation, reducing back-out and catheter complications. Advance on AP, then confirm depth on lateral; if you hit a spinous process, back out and adjust the angle. Free CSF backflow on removing the stylet confirms entry. (The same paramedian approach is useful for the test dose in patients with severe scoliosis, where a standard L4–L5 route is difficult.)
Set the catheter-tip level to the spasticity/pain pattern. Thread the catheter under fluoroscopy to the target level: for lower-limb-only spasticity the tip sits around the upper lumbar to low-thoracic cord (up to ~T6); for combined upper- and lower-limb involvement it is advanced higher, to about C7–T1. Adjust in or out under live imaging until the tip is where you want it.
Anchor to the deep fascia. Make a small incision around the needle entry, slightly cranial, and carry the dissection down to the lumbodorsal fascia — the fascia is exposed, not opened, since the needle and catheter already pass through it. Seat the manufacturer anchor over the catheter against the fascia and fix it with non-absorbable sutures. When removing the stylet and needle and tying the anchor, take care not to drag the catheter or shift the tip. Catheter problems — kinking, migration, disconnection — are the commonest hardware failures.
Fashion the abdominal pocket. Site it over four finger-breadths between the costal margin and the iliac crest, horizontal incision at the superior margin, marked before draping; in lumbar scoliosis place the pump on the convex side. Dissect superior-to-inferior with blunt/finger dissection to fit the pump. In thin or small patients the abdominal wall muscle is easily penetrated — a real risk of peritoneal entry, fistula, and infection — so stay superficial to the fascia.
Tunnel, connect, confirm flow. Pass a tunneling device from the spinal wound to the pocket, draw the pump-segment catheter back to the spinal side, and trim excess only at the spinal wound after confirming enough length remains abdominally; leave a 360° loop at each end (strain relief, and slack for a growing child). Connect the two catheter segments with the supplied connector and watch for CSF backflow to reach the abdominal side — this can take a few minutes and confirms patency and a leak-free connection. A small-gauge needle in the catheter access port aspirating CSF confirms the pump is in continuity with the thecal sac.
Seat, anchor, program, close. Fill the reservoir with the prepared drug, seat the pump in the correct orientation, and anchor it with non-absorbable sutures — all four corners, or two at opposite corners — to prevent rotation/flipping. Irrigate copiously; close in layers with absorbable sutures to a watertight result; the representative interrogates and primes the pump and the neurologist's starting dose/rate is set. (Vancomycin powder in the pocket is practiced by some without strong supporting evidence.)

Operative pearls Use a paramedian oblique needle path (entry ~L4 pedicle, ~30° to enter L2–L3), not a straight midline stick — it prevents catheter kink and gives fascial purchase. Match tip level to the limbs involved (~T6 for legs only; C7–T1 if arms too). Trim catheter length only at the spinal wound, and leave a 360° loop at both ends. Confirm CSF backflow all the way to the pump side before final connection. Anchor the pump (2 opposite or 4 corners) to stop rotation. Stay superficial to the abdominal fascia in thin patients to avoid peritoneal penetration. In lumbar scoliosis, lay the patient in lateral decubitus with the curve's concavity down (toward the table) and convexity up (toward the surgeon): this opens the convex side of the interlaminar window for the paramedian needle and faces the side on which the pump pocket will sit.

4.Pump Replacement and System Priming

Most pumps are replaced at battery end-of-service (roughly every 6–7 years for the programmable electronic pumps); the remainder are revised for device or catheter malfunction or for infection. An elective battery exchange should be scheduled ahead of depletion, planned months in advance, because letting the pump run to end-of-life risks abrupt cessation and withdrawal. The replacement is usually simpler than the original implant — frequently an abdominal-pocket procedure alone — but it carries its own dosing hazards.

Before surgery, confirm the exact drug, concentration, daily dose, and reservoir volume from the device record so the replacement pump is filled and programmed to reproduce the current regimen, and reconfirm those numbers before closure. Inspect the original pocket and scar for any sign of infection or skin erosion, and plan the incision to use the existing scar and reach the pump cleanly.

Decide pump-only exchange versus catheter revision early, because it changes the prep and positioning. Access the system and confirm catheter patency by aspirating CSF: if CSF returns freely the catheter is patent and only the pump need be exchanged (an abdominal field alone); if CSF cannot be aspirated or a catheter problem is suspected, plan a catheter revision, which requires the spinal exposure and the lateral positioning of the original implant. Swap the pump, irrigate the pocket copiously, secure meticulous hemostasis to limit hematoma and seroma, re-seat and re-anchor the new pump against rotation, and close.

Clearing old drug, and the priming-bolus trap When the catheter is retained, its dead space still holds drug at the old pump's concentration. Aspirate about 1 mL back through the catheter access port: this both confirms patency (free CSF return) and draws the residual concentrated drug out of the line. The priming step is where the scenarios diverge and where overdoses have occurred. A new, empty catheter must be primed with a priming bolus so drug traverses the dead space to the tip — without it the slow basal rate would take days to reach the cord and the patient would drift toward withdrawal. A retained catheter that still contains drug must not be given a priming bolus after it is reconnected to the new pump: driving that high-concentration column into the cord can cause a clinically significant or fatal overdose. Follow the device protocol, confirm CSF return before relying on the system, and account for catheter dead-space volume whenever the drug or concentration changes; if CSF cannot be aspirated, treat the catheter as non-patent and plan revision.

5.Perioperative Care

Preoperative. Confirm a positive trial, optimize and document the baseline (pain scores, or a spasticity scale and functional goals), screen and clear infection, review anticoagulation, and counsel the patient on realistic expectations and the refill commitment. Give weight-based intravenous antibiotic prophylaxis at induction, following implantable-device guidance: a first-generation cephalosporin such as cefazolin within 30–60 minutes of incision, with vancomycin (begun within 120 minutes) reserved for documented MRSA risk or beta-lactam allergy. For spasticity, set goals explicitly — whether the aim is easier care and comfort or improved active function changes the target dose.

Intraoperative. Coordinate with anesthesia on neuromuscular blockade, which is generally avoided during implantation so that residual paralysis does not confound assessment of the baclofen effect: once the system is primed, any hypotonia, weakness, excess sedation, or respiratory depression should be attributable to the drug rather than to a lingering blocker, and the patient's tone and motor responses remain observable on emergence. No formal guideline mandates a particular regimen, so this is a practice pattern rather than a rule. Sterility and hemostasis at both wounds, and clean fluoroscopic access, remain the other intraoperative priorities.

Postoperative — early. Check the wound every few days but do not expose it unnecessarily; a soaked or soiled dressing may signal a seroma or CSF leak, initially managed with a pressure dressing (an abdominal binder over the pocket reduces dead space and seroma) and observation. Escalate wound checks and consider prophylactic antibiotics if a superficial infection is suspected. Other predictable early issues are spinal headache from the dural puncture and early over- or under-dosing as the starting rate is titrated.

Postoperative — ongoing. The neurologist titrates dose and rate to response over follow-up. The reservoir is refilled as a simple outpatient procedure every ~4–6 months (interval depends on reservoir size, concentration, and daily dose), typically with the neurologist and a device representative who reprograms settings. Follow patients regularly, because key problems are easily missed by caregivers: in particular, return of spasticity is an important late sign that may indicate failed drug delivery — a catheter kink, migration, or disconnection, or pump malfunction — and warrants device interrogation and imaging. A device/drug identification card should travel with every patient. (Current device guidance does not support implantation in children under four years of age.)

6.Drug Concentrations and Reservoir Volumes

The implanted pump is filled with a commercially available (or compounded) concentration of the chosen drug, and the reservoir volume — commonly 20 mL or 40 mL for the widely used programmable pumps — together with the daily dose sets the refill interval. The guiding principle is to choose a concentration high enough that the reservoir lasts a practical interval between refills, but no higher than needed, since concentrated opioid solutions raise the risk of catheter-tip granuloma. The figures below are representative of approved formulations; always confirm against current product labeling, pharmacy compounding, and your pump's specifications.

Representative intrathecal drug concentrations and reservoir volumes. Concentrations reflect commonly available approved/compounded preparations; confirm against current labeling and institutional protocol before use.
Drug	Indication	Available concentrations	Typical reservoir
Morphine (preservative-free)	Pain	10 and 25 mg/mL (concentrated); compounded up to ~30 mg/mL	20 or 40 mL
Ziconotide	Pain	25 µg/mL and 100 µg/mL	20 or 40 mL
Baclofen	Spasticity	Trial: 50 µg/mL · Pump: 500, 1000, 2000 µg/mL	20 or 40 mL

How concentration, dose, and refill interval relate Daily dose (drug per day) divided into the reservoir (concentration × volume) gives the refill interval. Raising the concentration extends the interval but concentrates the solution at the catheter tip; for opioids that trade-off is balanced against granuloma risk, which is why the lowest effective concentration that gives a reasonable refill interval is preferred. For baclofen, the three pump concentrations (500/1000/2000 µg/mL) let the daily microgram dose be delivered in a practical reservoir refill interval as the dose climbs.

Two dosing emergencies to teach every covering clinician Overdose / inadvertent intrathecal bolus — from a programming error, refill misadventure, or pocket fill (drug injected into the pocket rather than the refill port) — produces drug-specific toxicity: respiratory depression and excess sedation with opioid; profound hypotonia, somnolence, and respiratory depression with baclofen. Withdrawal — from pump failure, an empty reservoir, catheter disconnection, or battery depletion — is the mirror image and, for baclofen, is the true emergency covered below.

7.Managing the Common Complications

Complications fall into two families: wound and hardware problems around the pocket and catheter, and drug problems from the delivered agent. The wound issues are usually managed conservatively; the drug emergencies are managed by drug and require recognition more than heroics.

Seroma. Fluid collection over the pocket from residual dead space is common and usually self-limited; manage with an abdominal binder or pressure dressing and observation. Avoid percutaneous aspiration unless the collection is large, symptomatic, or threatening the wound — and then only under strict sterility, since aspiration can seed an otherwise sterile collection and convert it into an infected, explant-requiring problem.
Infection. Superficial wound infection may settle with antibiotics, but a deep pocket infection, infected catheter track, or meningitis almost always requires explantation of the entire system; partial salvage rarely succeeds. Watch for pocket erythema or breakdown, persistent fever, and any meningitic picture. Prevention is the better lever — perioperative antibiotics, meticulous sterility, and short operative time.
Overdose / inadvertent intrathecal bolus. Treat by drug: opioid overdose needs airway support and naloxone; baclofen overdose needs supportive care up to and including intubation for respiratory depression. In either case interrogate and stop the pump, and aspirating CSF through the access port can remove undelivered drug.
Baclofen withdrawal. The can't-miss emergency, beginning within 24–72 h of an interruption — rebound spasticity, fever, rigidity, and altered mental status that mimics sepsis or neuroleptic malignant syndrome. Restore intrathecal delivery (correct the cause; temporary intrathecal baclofen if needed) and bridge multimodally with enteral baclofen, benzodiazepines, and cyproheptadine. Detailed, with the device workup, in the spasticity section below.
Catheter-tip granuloma. A sterile inflammatory mass at the catheter tip (detailed in the box below) — suspect it when analgesia fades and doses climb without true tolerance, or when new neurological signs appear.

Catheter-tip granuloma — the board-favorite, in depth The catheter-tip inflammatory mass is a sterile (non-infectious) granuloma arising from the arachnoid, usually intradural-extramedullary. It is dose- and concentration-dependent and most strongly associated with morphine (also hydromorphone and other opioids; ziconotide carries little granuloma risk, which is part of why it is a useful switch). The cardinal clue is declining analgesia that needs escalating doses but is not explained by tolerance — failing delivery from the enlarging mass — sometimes with mass effect: new radicular pain at the catheter-tip level, or myelopathy. It develops slowly, typically over months to a couple of years (average time to diagnosis around two years). Diagnose with MRI with and without contrast (a rim-enhancing mass) or CT myelography; reported incidence is low (well under 1% in survey data, higher in some series).

Management turns on symptoms. For an asymptomatic (often incidental) granuloma: lower the drug concentration and dose, favor bolus over continuous dosing, switch toward ziconotide or another agent, and/or withdraw the catheter a few centimeters — stopping or reducing the opioid usually lets the mass involute. For a symptomatic mass with cord compression or a progressive deficit, the board answer is surgical: do not simply pull the catheter back — perform open decompression, and do not chase a gross-total resection if the mass is adherent to the cord.

On concentration specifically: risk rises with high concentration and dose, classically with compounded morphine at up to ~50 mg/mL used to stretch refill intervals. The commercially available preparations are 10 and 25 mg/mL, and consensus (PACC) ceilings are about 20 mg/mL concentration and 15 mg/day dose for morphine; when a granuloma appears, reduce toward the lowest effective concentration within those limits. The caveat worth remembering is that no concentration is fully granuloma-sparing — masses have formed even at low dose and concentration — so reducing concentration lowers, but does not abolish, the risk. Prevention remains the recurring discipline of intrathecal opioid therapy: the lowest effective concentration and dose.

Part II

Intrathecal Analgesia

8.Cancer Pain

The clearest role for intrathecal analgesia is refractory cancer pain — pain inadequately controlled despite aggressive systemic opioids, or controlled only at the cost of intolerable toxicity (sedation, confusion, nausea, constipation). Here intrathecal delivery does two things at once: it improves analgesia and it strips away the systemic drug burden, often returning a clear sensorium to a patient who had been obtunded by escalating oral or IV opioids.

The pivotal randomized trial (Smith and colleagues) compared an implantable intrathecal drug delivery system plus comprehensive medical management against medical management alone in refractory cancer pain. The IDDS arm achieved better pain and toxicity outcomes — more patients met the combined endpoint of reduced pain and reduced drug toxicity — and the trial is widely cited for the additional observation of better survival in the IDDS group, a finding usually attributed to reduced systemic opioid toxicity. The clinical translation is simple: in the patient drowning in systemic opioids, moving the drug intrathecally can restore both comfort and lucidity. This is the non-destructive sibling of cordotomy and myelotomy, and it is often the better choice for diffuse, midline, or bilateral pain that a unilateral lesion could not cover.

9.Chronic Non-Cancer Pain and the Drugs

In carefully selected chronic non-cancer pain, intrathecal therapy is a longer-term commitment with a higher bar for selection (including psychological evaluation), reserved for patients who have failed less invasive measures and demonstrated benefit on trial. The two FDA-approved intrathecal analgesics anchor the regimen:

Preservative-free morphine — the workhorse opioid; effective but carries the opioid liabilities of tolerance, dose escalation, respiratory depression (especially at initiation), and catheter-tip granuloma at higher concentrations.
Ziconotide — a non-opioid N-type calcium-channel blocker; no respiratory depression and no tolerance, but a narrow therapeutic window and neuropsychiatric side effects that demand slow titration.

Beyond these, admixtures (e.g., local anesthetic, clonidine) are used per consensus guidance. The catheter-tip inflammatory granuloma — a mass that can compress the cord — is the characteristic long-term opioid complication to keep in mind, presenting as loss of efficacy or new neurological signs.

Part III

Intrathecal Baclofen for Spasticity

10.Who Benefits

Intrathecal baclofen (ITB) treats severe spasticity that is diffuse, that interferes with care, comfort, positioning, or function, and that cannot be controlled by oral agents without intolerable sedation. The classic populations are spasticity of cerebral origin (cerebral palsy, acquired brain injury, stroke) and of spinal origin (spinal cord injury, multiple sclerosis). Baclofen, a GABA-B agonist, acts at the spinal cord; delivering it intrathecally achieves at the cord what oral dosing cannot, while sparing the patient the sedation of high oral doses. As with pain, candidacy is confirmed by a test-dose trial showing a clear, dose-related fall in tone, and goals are set explicitly — reduced tone to ease caregiving and prevent contracture is a different target from improving volitional movement.

Tone is graded with the modified Ashworth scale, the standard bedside measure for both selection and follow-up; a fall of at least one grade after the test dose is the conventional marker of a positive trial.

Modified Ashworth Scale for grading resistance to passive movement.
Grade	Description
0	No increase in muscle tone.
1	Slight increase in tone — a catch and release, or minimal resistance at the end of range of motion.
1+	Slight increase in tone — a catch, followed by minimal resistance through the remainder (less than half) of the range.
2	More marked increase in tone through most of the range, but the affected part is still easily moved.
3	Considerable increase in tone; passive movement is difficult.
4	Affected part rigid in flexion or extension.

Intrathecal baclofen withdrawal — the can't-miss emergency Abrupt interruption of intrathecal baclofen — empty reservoir, catheter disconnection or migration, pump failure, battery end-of-life, or a missed refill — produces a withdrawal syndrome that can be life-threatening. It typically begins within 24–72 hours of the interruption and presents with rebound spasticity often worse than baseline, high fever, altered mental status, tachycardia, and labile blood pressure, progressing in severe cases to rhabdomyolysis, multi-organ failure, and death. It is frequently mistaken for sepsis, neuroleptic malignant syndrome, malignant hyperthermia, or autonomic dysreflexia. The definitive treatment is restoring intrathecal delivery — correct the pump/catheter cause, and where needed give a temporary intrathecal (lumbar) baclofen bolus or infusion. Because oral baclofen barely reaches the CSF (even five- to ten-fold oral doses deliver little to the cord, so it is at best a weak adjunct, not a substitute for intrathecal drug), bridge multimodally: escalating benzodiazepines are the practical mainstay, with cyproheptadine (a serotonin antagonist, ~4–8 mg every 6–8 h), dantrolene or tizanidine for rigidity, and a propofol infusion for refractory cases in the ICU, all under aggressive supportive care and active cooling for hyperthermia. Any ITB patient who presents febrile, rigid, and confused is in withdrawal until proven otherwise — do not wait for the sepsis workup to return.

Finding the cause: working up a suspected delivery failure Interrogate the pump first — telemetry confirms it is running, the programmed rate, reservoir volume, and battery status, and can reveal a stalled rotor. Plain AP and lateral radiographs of the entire pump-catheter system show disconnection, fracture, kinking, or migration. The decisive test is the catheter access (side) port study: aspirate through the port under fluoroscopy — if CSF cannot be aspirated, do not inject, since a non-patent or malpositioned catheter risks driving an inadvertent bolus; if CSF returns, first withdraw about 1 mL to clear the residual concentrated drug from the catheter dead space, then inject contrast to confirm intrathecal flow, often with CT — injecting without aspirating first can push that drug bolus into the cord and cause overdose. A radionuclide (indium-111) study can localize a slow or occult leak. One pitfall: initial CSF return does not fully exclude an extradural catheter tip, so correlate aspiration with the contrast study. Correcting the cause is usually surgical — pump replacement or catheter revision — the main exception being a simply exhausted reservoir, which a refill fixes; when the catheter is correctly positioned but the pump has failed, baclofen can be given directly through the access port as a temporizing measure.

11.Comparison at a Glance

The two intrathecal therapies share hardware and operation but differ in everything clinical. Doses are representative starting points and must be confirmed against current device labeling and institutional protocol.
	Intrathecal analgesia	Intrathecal baclofen (ITB)
Indication	Refractory cancer pain; selected chronic non-cancer pain	Severe spasticity (cerebral or spinal origin)
Drugs	Preservative-free morphine; ziconotide; admixtures	Baclofen (GABA-B agonist)
Trial	IT bolus or short infusion; ≥~50% pain relief	Test dose ~50–100 µg; fall in Ashworth tone
Catheter tip	Spinal level of the painful region	Higher (often mid/low thoracic; higher for arm tone)
Signature long-term issue	Tolerance; catheter-tip granuloma (opioid)	Withdrawal emergency; dose drift
Overdose picture	Respiratory depression, sedation (opioid)	Hypotonia, somnolence, respiratory depression

Part IV

Pearls

One system, two therapies: intrathecal analgesia (cancer and selected chronic pain) and intrathecal baclofen for spasticity share pump and catheter but differ in drug, trial, target level, and emergencies.
The principle is dose economy: drug at its spinal site of action means a small intrathecal dose replaces a large, toxic systemic one.
Trial before you implant (except in short-prognosis cancer): a test dose or short infusion proves benefit in this patient before committing to a pump.
For cancer pain, IDDS improves both pain and toxicity and is the non-destructive alternative to cordotomy/myelotomy — especially for diffuse, midline, or bilateral pain.
Two approved intrathecal analgesics: preservative-free morphine (effective, but tolerance, respiratory depression, granuloma) and ziconotide (no respiratory depression or tolerance, but narrow window and neuropsychiatric effects).
Catheter-tip granuloma presents as lost efficacy or new neurologic signs in a long-term opioid patient — image it.
Anchoring and tunneling technique decide durability: catheter migration, kinking, and disconnection are the commonest hardware failures.
Operative specifics: paramedian oblique needle (entry ~L4 pedicle, ~30° into L2–L3) to avoid kink; tip ~T6 for legs only or C7–T1 if arms involved; trim catheter only at the spinal wound and leave a 360° loop both ends; confirm CSF backflow to the pump side before connecting; anchor the pump to prevent rotation.
Return of spasticity in a stable ITB patient is a delivery-failure sign — suspect catheter kink/migration/disconnection or pump malfunction and interrogate/image; do not just assume tolerance.
Baclofen test dose: ~50 µg by LP, injected slowly over ~1 min; assess at peak effect ~4 h later for a ≥1-point modified Ashworth drop; escalate to 75/100 µg on later days if no response.
Don't abolish useful tone: if a patient relies on extensor spasticity to stand or transfer, ITB may leave them functionally worse — weigh this before trialing.
Distinguish spasticity from fixed contracture: baclofen relaxes dynamic tone but cannot release an established joint contracture or arthrosis — a limb that gains no range on the test dose signals a contracture to address surgically/orthopedically, not with a pump.
Intrathecal baclofen withdrawal is the can't-miss emergency, onset within 24–72 h: fever, rigidity, rebound spasticity, and altered mentation that mimics sepsis/NMS. Oral baclofen penetrates the CNS too slowly to rely on — restore intrathecal delivery and bridge with benzodiazepines and cyproheptadine; do not wait for the sepsis workup.
Work up a suspected delivery failure systematically: pump interrogation, then AP/lateral radiographs, then a catheter access-port study — aspirate under fluoroscopy and never inject contrast if CSF cannot be withdrawn.
Every pump patient should carry device and drug identification — emergency management depends on knowing the drug, concentration, and rate.

Selected References

Smith TJ, Staats PS, Deer T, et al. Randomized clinical trial of an implantable drug delivery system compared with comprehensive medical management for refractory cancer pain: impact on pain, drug-related toxicity, and survival. J Clin Oncol. 2002;20(19):4040–4049. The pivotal IDDS-for-cancer-pain RCT.
Deer TR, Pope JE, Hayek SM, et al. The Polyanalgesic Consensus Conference (PACC): recommendations for intrathecal drug delivery: guidance for improving safety and mitigating risks. Neuromodulation. 2017;20(2):96–132. Consensus framework for drugs, trialing, and safety.
Deer TR, Prager J, Levy R, et al. Polyanalgesic Consensus Conference 2012: consensus on diagnosis, detection, and treatment of catheter-tip granulomas (inflammatory masses). Neuromodulation. 2012;15(5):483–495. Diagnosis and stepwise management of catheter-tip granuloma.
Penn RD, Savoy SM, Corcos D, et al. Intrathecal baclofen for severe spinal spasticity. N Engl J Med. 1989;320(23):1517–1521. The foundational ITB trial.
Coffey RJ, Edgar TS, Francisco GE, et al. Abrupt withdrawal from intrathecal baclofen: recognition and management of a potentially life-threatening syndrome. Arch Phys Med Rehabil. 2002;83(6):735–741. The definitive description of the withdrawal emergency.
Meythaler JM, Roper JF, Brunner RC. Cyproheptadine for intrathecal baclofen withdrawal. Arch Phys Med Rehabil. 2003;84(5):638–642. Case series supporting cyproheptadine plus benzodiazepine and baclofen in the multimodal bridge.
Miracle AC, Fox MA, Ayyangar RN, et al. Imaging evaluation of intrathecal baclofen pump-catheter systems. AJNR Am J Neuroradiol. 2011;32(7):1158–1164. Radiography, catheter-access-port contrast study, and scintigraphy for suspected delivery failure.
Rauck RL, Wallace MS, Leong MS, et al. A randomized, double-blind, placebo-controlled study of intrathecal ziconotide in adults with severe chronic pain. J Pain Symptom Manage. 2006;31(5):393–406. Pivotal ziconotide evidence.
Hayek SM, Hanes MC. Intrathecal therapy for chronic pain: current trends and future needs. Curr Pain Headache Rep. 2014;18(1):388. Practical overview of patient selection and management.