What Is Intrathecal Administration and How Does It Work?

Intrathecal administration is the delivery of medication directly into the fluid-filled space surrounding the spinal cord and brain, bypassing the bloodstream entirely. A needle or catheter passes through the membranes that encase the central nervous system, depositing drugs into the cerebrospinal fluid (CSF) so they can act on nerve tissue without first traveling through the rest of the body. This route is used for surgical anesthesia, chronic pain management, spasticity treatment, and a growing list of neurological conditions, and the way drugs behave once they reach that narrow space is more complex than it might seem.

Where the Drug Actually Goes

The spinal cord sits inside a layered set of protective membranes. The outermost is the dura mater, a tough sheath. Just beneath it lies the arachnoid membrane, which regulates what diffuses inward. Together, these two layers form the dural sac. Inside that sac, between the arachnoid and the delicate pia mater that directly covers the spinal cord, is the subarachnoid space. This is where cerebrospinal fluid circulates, and it is the target of an intrathecal injection. The pia mater itself is permeable, allowing drugs to pass through gaps between its cells and reach spinal cord tissue underneath.¹

The practical upshot is that a clinician advances a needle through skin, muscle, and those two outer membranes to reach the CSF. Once a drug is deposited there, it sits in direct contact with the central nervous system. That is a dramatically shorter path to its target than an oral pill or intravenous drip would take, and it means far smaller doses can produce powerful effects.

Why Not Just Use a Pill or an IV

The central nervous system is guarded by the blood-brain barrier, a tightly sealed layer of cells lining the blood vessels in the brain and spinal cord. Most molecules in the bloodstream cannot cross it. For many neurological conditions, this barrier is the core problem: a drug that works beautifully in a lab dish never reaches the tissue it needs to treat because the barrier blocks it. Intrathecal delivery sidesteps that bottleneck. It also avoids the systemic exposure that causes side effects elsewhere in the body, making it possible to use drugs that would be too toxic if given intravenously.²

Opioids are a clear example. When morphine is given intrathecally for postoperative pain, doses as low as 100 micrograms can produce intense pain relief lasting up to 24 hours. That is a tiny fraction of what would be needed orally, which means fewer side effects like sedation, nausea, and constipation.³

How Drugs Spread Through the CSF

Once injected, a drug does not just sit at the injection site. It disperses through the cerebrospinal fluid, and this happens much faster than simple diffusion would predict. The reason is that CSF is not still. It pulses rhythmically, driven by the heartbeat. Each cardiac cycle pushes a small wave of fluid up and down the spinal canal, and these oscillations stir the drug along far more efficiently than molecular diffusion alone.⁴

The anatomy inside the subarachnoid space matters, too. Tiny structures like nerve roots, ligaments, and blood vessels create turbulence that further enhances mixing. Computational modeling has shown that these small anatomic features can increase how far a drug spreads along the spine by roughly 60% compared to a smooth, featureless canal. Where exactly the catheter tip sits, its angle, and how fast the drug is injected also shift the distribution pattern substantially.⁵

This matters clinically because the goal is usually to concentrate the drug near a specific segment of the spinal cord, say the lumbar region for leg pain. Factors like patient position, injection speed, and even the density of the drug solution relative to CSF all influence where the medication ends up. Getting this wrong can mean inadequate pain relief in the target area or unwanted effects higher up the spinal cord.

New barriers also come into play once the drug is in the CSF. The fluid itself turns over regularly as the body produces and reabsorbs it, which means drugs get cleared before they can fully penetrate tissue. Reaching deep brain targets from a lumbar injection is especially difficult because of this constant flushing.⁶

Chronic Pain and the Intrathecal Pump

For people with severe chronic pain that does not respond to conventional treatments, an implanted intrathecal drug delivery system can be a long-term option. A small pump, roughly the size of a hockey puck, is surgically placed under the skin of the abdomen. A thin catheter tunnels from the pump to the intrathecal space. The pump delivers a precisely programmed flow of medication directly into the CSF, day and night, and can be refilled periodically through the skin with a needle.

Morphine is the most commonly used drug in these pumps, but it is not the only one. Ziconotide, a non-opioid pain drug derived from cone snail venom, has shown strong results in patients with refractory cancer pain or AIDS-related pain. In a randomized controlled trial, patients receiving intrathecal ziconotide saw their pain scores improve by about 53% on average during the initial dosing phase, compared with about 18% in the placebo group. Half of the patients on ziconotide responded to therapy, and five achieved complete pain relief.⁷ In clinical practice, ziconotide is sometimes added to intrathecal morphine when opioid therapy alone is not enough, with pain reduction typically appearing within a few days of starting the combination.⁸

Before someone gets an implanted pump, they usually undergo a trial. This involves a temporary intrathecal injection, either a single bolus or a short continuous infusion through a temporary catheter, to see whether the drug gives meaningful relief. Interestingly, research comparing these two trialing methods found no significant difference in their ability to predict who would benefit from a permanent implant.⁹ For ziconotide trials, a response is generally defined as at least a 30% reduction in pain with no intolerable side effects on at least two occasions.¹⁰

One concern with long-term intrathecal opioid therapy is tolerance and dose escalation. A small controlled trial that reduced intrathecal morphine doses by 20% per week in chronic non-cancer pain patients had to stop recruitment early because the dose-reduction group deteriorated significantly, with pain scores climbing sharply, while the stable-dose group held steady. That finding underscores how physically dependent the nervous system can become on a continuous intrathecal opioid supply.¹¹

Treating Severe Spasticity

Baclofen, a muscle relaxant, is another drug frequently given intrathecally through an implanted pump. Oral baclofen helps many people with spasticity from conditions like cerebral palsy, multiple sclerosis, or spinal cord injury, but some patients cannot tolerate the high oral doses needed, or the drug simply does not reach the spinal cord in sufficient concentration. An intrathecal baclofen pump solves both problems by putting the drug exactly where it acts.

Evidence supports its effectiveness for both short-term and long-term reduction of severe spasticity in patients who have failed oral treatment, with additional qualitative evidence showing functional improvement in daily activities.¹² The challenge, as a recent narrative review emphasized, is optimizing dosing to balance efficacy against safety risks, because baclofen withdrawal from a pump malfunction can cause life-threatening symptoms including high fever, altered consciousness, and rebound spasticity.¹³

Spinal Anesthesia for Surgery

The most familiar use of intrathecal injection is probably spinal anesthesia for surgery, particularly cesarean deliveries and lower-body procedures. A single injection of a local anesthetic like bupivacaine into the lumbar intrathecal space produces rapid, dense numbness from the mid-torso down. Whether the anesthetic solution is formulated to be heavier than CSF (hyperbaric) or the same density (isobaric) affects how it spreads once injected. A Cochrane review of trials in cesarean sections found that hyperbaric bupivacaine produced a faster onset of the sensory block and was less likely to require conversion to general anesthesia compared with isobaric bupivacaine.¹⁴

Heavier solutions sink with gravity, which makes the spread more predictable when the patient is positioned a certain way. Isobaric solutions stay closer to where they are injected, which can be an advantage when targeted anesthesia at a specific spinal level is the goal but a disadvantage if the block ends up too patchy.

Needle Design and Headache Risk

One of the most common complaints after any lumbar puncture or spinal anesthetic is a postdural puncture headache, a distinctive headache that worsens when sitting or standing and improves when lying flat. It happens because the hole left in the dura allows CSF to leak out, lowering the fluid pressure around the brain. The type of needle used makes a real difference. A meta-analysis of 25 randomized trials found that pencil-point needles, which spread rather than cut the dural fibers, produced significantly fewer headaches than traditional cutting-tip needles. Severe headaches and the need for an epidural blood patch to seal the leak were also much less common with pencil-point designs.¹⁵

Complications of Implanted Systems

Implanted intrathecal pumps are generally safe, but the hardware sits inside the body for years, and complications do occur. An analysis of 282 intrathecal pump operations, split between baclofen and opioid therapy, found that about one in five patients developed a CSF leak, and roughly 5% developed a deep infection. Risk factors for leaks included tobacco use and younger age, while infection was associated with being underweight, immunosuppression, tobacco use, and being a first-time implant rather than a revision.¹⁶

A rarer but more serious complication is the formation of a granuloma, a small mass of inflammatory tissue, at the tip of the catheter. Catheter-tip granulomas can compress the spinal cord and cause neurological deficits if they grow large enough.¹⁷ A systematic review of reported cases found that the morphine dose and concentration delivered through the catheter were strongly correlated with granuloma development, suggesting that higher opioid concentrations at the catheter tip increase the risk.¹⁸ Clinicians monitor for new or changing neurological symptoms, and management may involve switching away from morphine, lowering the concentration, or surgically repositioning the catheter.

What Goes Into the Pump Matters

Because implanted pumps hold medication at body temperature for weeks at a time, the chemical stability of the drug mixture inside the reservoir is a real concern. Many patients receive combinations of two or even three drugs in a single admixture. Testing has shown that a combination of morphine, bupivacaine, and clonidine remains stable for at least 90 days at body temperature inside a pump, with all three drugs maintaining above 96% of their starting concentration.¹⁹ A morphine-clonidine combination similarly held above 94% over three months.²⁰

Not all combinations are equally forgiving. When ziconotide is mixed with commercially formulated baclofen and stored at body temperature, the ziconotide concentration drops to about 82% of its initial value within 30 days, reaching the 90% stability mark at only 12 days. Using powdered baclofen instead of the commercial liquid formulation extends that 90% stability window to about 20 days.²¹ This means the pump has to be refilled more often when ziconotide is in the mix, and pharmacists need to choose their formulation ingredients carefully.

Gene Therapy and Newer Biologics

Intrathecal delivery has taken on a new dimension with the arrival of gene therapies for neurological diseases. Onasemnogene abeparvovec, the gene therapy for spinal muscular atrophy, was originally given intravenously. A phase 3b trial tested intrathecal delivery of the same therapy in patients who had already received prior treatment, including those already on nusinersen (another drug given intrathecally). All 27 participants experienced at least one side effect, most commonly upper respiratory infections, fever, and vomiting. Motor function scores showed modest and variable changes over 52 weeks.²² The intrathecal route is attractive for gene therapy because it requires far less viral vector than an IV infusion, potentially reducing the immune response and liver toxicity that high systemic doses can cause.

Challenges in Children

Intrathecal therapy is not simply adult therapy scaled down. In children, the volume of CSF does not increase in a straight line with age or body size, making dose calculations tricky. A drug concentration that is appropriate for a teenager may be too high for a small child or too low for an adolescent approaching adult size. The rate at which drugs are cleared from the CSF may also differ in pediatric patients.²³ Intrathecal chemotherapy for childhood leukemia is one of the oldest pediatric uses of this route, and dosing protocols typically rely on age-based rather than weight-based schedules precisely because CSF volume correlates better with age in the growing years.

Where the Technology Is Heading

One limitation of current intrathecal therapy is that injected drugs wash away as CSF turns over, requiring either repeated injections or a permanent pump. Researchers are working on injectable hydrogels, gel-like materials that can be placed in the intrathecal space and slowly release drug over days or weeks. One approach uses a biopolymer matrix made from hyaluronan and methylcellulose loaded with drug-carrying nanoparticles. In animal models of spinal cord injury, this system delivered medication locally to the injured cord while bypassing the blood-spinal cord barrier.²⁴ The same platform has been adapted to carry growth factors like neurotrophin-3, maintaining the protein’s biological activity during sustained release.²⁵

These depot systems are still in early-stage research, but they address a real gap. A single injection that provides weeks of local drug release would reduce the number of procedures a patient needs, lower infection risk, and potentially improve outcomes for conditions like spinal cord injury where consistent drug exposure matters. Combined with the expanding use of intrathecal gene therapy and biologics, the space is moving well beyond the morphine pumps and spinal anesthetics that defined it for decades.

A Brief Origin Story

Intrathecal access has been a medical tool for well over a century. In the 1870s, Heinrich Quincke in Germany began experimenting with lumbar puncture in animals, and by 1891 he had published a series of lumbar punctures performed on patients with tubercular meningitis to relieve elevated pressure inside the skull. He designed the spinal needle that made reliable access possible. August Bier took the next step in 1898, performing what is generally credited as the first spinal anesthetic: he injected small doses of cocaine into the intrathecal space of six surgical patients, publishing his results the following year.²⁶ The route that started with a needle and a vial of cocaine now delivers everything from programmable opioid infusions to gene-replacing viral vectors, but the fundamental idea remains the same: put the drug where it needs to be, and let the body’s own barriers work for you instead of against you.