Drugs spread within the CSF spread via several mechanisms including rate and volume of injection, baricity of the injectate, and hydrophilicity. Infusions of drugs do not spread far beyond the area of injection. The drugs stay within the ring in which it is injected [18–20]. Infusions with animal models showed that the dispersion of fluid from the catheter tip was minimal. This was shown with intrathecal infusions performed with methylene blue in pig models. The spread of the dye infused through a catheter was observed and showed that it did not spread more than four vertebral levels above or below the level of the catheter tip [20]. This is because of the relatively stagnant flow of CSF within the spinal canal. The CSF rather has regions of intermixing and stays within a “bubble” or “ring” [21–24]. On the other hand, bolus injections are distributed more widely because they go beyond the ring of an infusion. This is because bolus injections are generally a larger volume injected at a faster rate when compared to an infusion. Properties of the injectate also play a large role in the spread of the injectate. Assuming an upright sitting patient, the more hyperbaric a solution is the more caudad the solution will diffuse. The more hypobaric a solution is the more cephalid the solution will diffuse. Lipid solubility of a drug also affects the extent to which an injectate will diffuse. The more lipophilic (lipid soluble) a solution the less it will diffuse within the CSF. Lipophilic drugs are also more rapidly absorbed and subsequently excreted. The opposite is true of hydrophilic (water soluble) solutions that will spread more widely within the CSF and are less tissue soluble. This leads to a longer duration of action for hydrophilic solution. Hence, morphine (hydrophilic) has a longer duration of action than fentanyl (lipophilic). The size and polarity/ionization of a drug will also impact the spread of a drug. For example, ziconotide which is ionized, large molecule will have more diffuse spread within the CSF than most opioids that are generally small and unionized molecules.

Prior models of the CSF distribution proposed that CSF flow followed a rostral to caudal and subsequently caudal to rostral pattern. This was based off the discovery that the ventricles produced the CSF and resulted in the flow of CSF. This, however, recently has been disputed with a new theory that the CSF flows in a pulsatile fashion. At the level of the spine, the CSF oscillates in relation to the heart rate with areas of turbulence around the exiting nerve roots and walls of the spinal canal. This is significant and better explains why medications within the spinal CSF and ventricular CSF do not travel readily.

Intrathecal pumps use these principles to treat patients as an alternative to traditional therapy. The success of pumps and the medications used rely on the physiological principles described as earlier. We will discuss the mechanisms of action for specific medications in later subsections.

The IDDS consists of a small battery-powered pump/chamber and a spinal catheter that are implanted surgically. The chamber contains a drug or combination of drugs set by the prescriber that are pumped into the intrathecal space via the spinal catheter. The pumps can infuse at constant rates or programmable rates. A constant rate would be prescribed by the physician based on the patient’s needs using a radiofrequency transmitter handheld device called a physician programmer. With the constant rate of drug delivery, the assumption is that the patient’s drug requirement does not change significantly throughout the course of the day or night.

Programmable rates are variable and set by the prescriber using a physician programmer. The rates are based on the patient’s requirements, drugs administered, and concentration of drugs. Variable rates are useful for patients that have requirements that change during the day and/or are much less at night. This can avoid over-administration of drugs and allow for better patient satisfaction. Additionally, some pump systems like Synchromed II by Medtronic, allow for patients to self-administer medication bolus doses within physician prescribed limits. This is generally accomplished with the use of a hand-held device. This allows the patient to take more control over their analgesia and ultimately leads to greater patient satisfaction.

The success of any therapeutic modality depends a great deal on selecting patients appropriate for the intervention. The decision to proceed with intervention can be as important as forgoing the intervention and treating the patient via conservative means. Health care providers use guidelines, indications, risks, and benefits to decide if and which therapeutic intervention would be best. The Polyanalgesic Consensus Committee (PACC) in 2012 and the British Pain Society (BPS) in 2008 made the most recent recommendations for the selection and implantation of IDDS. Unfortunately, there are currently no universally accepted guidelines for the use of IDDS. We will use the PACC and BPS recommendations and systematic reviews to discuss patient selection, indications, and expert recommendations. For the purposes of patient selection, we simplify categorizing patients into two categories.

According to the World Health Organization’s (WHO’s) pain relief ladder for patients with cancer, patients are initially treated with nonopioids for mild pain, nonopioids and adjuvants for mild-to-moderate pain and finally opioids, nonopioids and adjuvants for moderate to severe pain [25]. Nonopioids analgesics generally refer to NSAIDs and acetaminophen. Adjuvants analgesics refer to classes of drugs that fall within the antiepileptics or antidepressants class. Conservative treatment generally entails using oral pain medications as listed within the pain relief ladder as well as less invasive therapy. Within the pain community, it is well understood that all pain patients benefit from a multidisciplinary approach from pain physicians, physiatrists, physical medicine and rehabilitation, and psychiatry. In addition to the WHO pain relief model, cancer pain patients may have benefit from other treatment modalities that include but are not limited to physical therapy, behavioral counseling/therapy, relaxation techniques, support groups, and acupuncture (Fig. 12.1).

As a patient’s pain progresses with disease pathology, the prescriber can move up the pain relief ladder to tailor the patient’s therapy. When an opioid reaches a point where it is no longer producing appreciable analgesia secondary to tolerance or intolerable side effects with increased dosing, then the health care provider should “rotate” the opioids. Opioid rotation is a vital part of the process as some opioids may respond better with certain patients and may not elicit unwanted side effects. Each patient’s response to an opioid is not predictable as responses may vary with different opioids. Before escalating medication dosages, the opioids should be “rotated” or changed. Once a medication reaches a ceiling point where the pain relief is not elicited or is intolerable due to side effects and other forms of conservative therapy have failed, intrathecal pumps can be considered.

Traditionally, patients with a prognosis of 3 months or less were not candidates for IDDS. In 2012, PACC revised their previous recommendations and added that patients with a prognosis of 3 months or less can be considered for intrathecal drug therapy [26]. Deer et al. showed that life expectancy can be improved with intrathecal drug therapy and if a patient is not imminently terminal, then they could be a candidate for implantable IDDS [26]. Additionally, IDDS therapy can be useful for patients undergoing chemotherapy/radiation. The first and only randomized controlled trial (RCT) completed to date examining implantable IDDS therapy was done on patients with cancer. In this trial, Smith et al. examined 202 patients that were randomly allocated into conventional medication management treated with morphine and interventional IDDS therapy with morphine infusions. The outcome showed that the IDDS therapy improved pain control, reduced systemic toxicity, and improved survival of patients with cancer [5]. There was a Cochrane review in 2005 evaluating the treatment of cancer pain intrathecal morphine compared to traditional routes. Twenty-eight cohort studies examined opioid intrathecal therapy and showed an 85 % success rate versus a 71 % success rate with traditional morphine therapy [27]. Unfortunately, there is still a paucity of strong literature supporting intrathecal drug therapy. The evidence is level II-C for IDDS therapy for patients with cancer, which means that the evidence is weak and evidence “suggestive” of using IDDS for patients with cancer [28]. The consensus among both PACC and BPS is that patients with cancer that may benefit from IT therapy and should be considered.

As with cancer pain, refractory noncancer or nonmalignant pain can be treated with IDDS. Some of the more common indications for IDDS are failed back syndrome (postlaminectomy syndrome), spinal stenosis, spondylosis, spondylolisthesis, complex regional pain syndrome (CRPS/RSD), neuropathies, spinal cord related injury, chronic pancreatitis, spasticity, and rheumatoid arthritis [26].

For patients with noncancer or nonmalignant pain, the selection process for proceeding with intrathecal pump implantation can become more complex. Much like the WHO pain relief ladder for cancer pain, the same principles can be applied to other pain patients. Additionally, if a conservative therapy fails (i.e., NSAIDs, opioids, adjuvants, physical therapy, biofeedback, etc.), they can be considered for implantable IDDS [26].

The literature for use of IDDS systems for chronic nonmalignant pain patients is less robust than the literature for patients with cancer. Patel et al. reported level II to level III evidence in favor of using IDDS therapy for chronic nonmalignant pain [29]. Moreover, in this systematic review, no randomized controlled trials met the inclusion criteria. Thus far, we have observational and open trial studies that can limitedly support the use of chronic nonmalignant pain.

For spasticity, the literature is slightly more robust for intrathecal therapy. This was shown in a 2006 Cochrane review where Tarrico et al. concluded that intrathecal baclofen produced significant analgesia for patients suffering from refractory spasticity [30]. Two double-blinded studies on spinal cord injury patients suffering from spasticity showed that intrathecal baclofen therapy was beneficial and did not cause significant side effects when compared to placebo [31, 32].

Pain can be a manifestation of physical, psychological, or behavioral factors. When treating pain, the health care provider must be cognizant of what can affect pain. The International Association for the Study of Pain defines pain as an “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage”[33]. Therefore, when considering an intrathecal pain pump, psychological or behavioral sources of pain should be evaluated and treated prior to implanting a pump. The psychological pain and stressors are important to address with terminal patients. Much of their pain and suffering can be treated with behavioral or psychological interventions [26]. Coping with the stressors of impending mortality, disability, and pain which afflicts nearly 30–50 % of patients with cancer. The benefit of analgesia outweighs the risk of untreated psychological or behavioral factors that could impact pain perception. Therefore, it is recommended to get psychological screening for IDDS candidates with nonmalignant pain and malignant pain.

Exceptions can be made in select scenarios. For example, in the case of patients with cancer where the prognosis may not be entirely clear, delaying implantable intrathecal therapy for psychological testing or even trialing for that matter may be time consuming and may delay significant analgesia that is much needed. Another instance where exceptions can be made is with stroke or cerebral palsy patients who suffer from spasticity. It is reasonable to continue to move forward with a trial/implant if psychological screening would be unproductive [26].

Expert panel recommendations as well as insurance companies will recommend psychological screening and treatment prior to trial and implantation as is seen with spinal cord stimulator trials. The goal of the screening is to help the patient understand the goal and treatment that will take place. It is to make reinforce the expectations from the intervention.

One of the main indications for intrathecal pump use is intractable pain that has failed conservative treatment. Failure of conservative treatment can be from ineffectiveness of previously effective medications due to tolerance or medication intolerance due to unwanted side effects. Approximately 10–30 % of chronic pain patients do not respond to conservative traditional oral treatments [34].

Another indication to proceeding with an intrathecal pump is the successful use of a medication for analgesia. This may seem fairly obvious, but it is important to know. The lack of success of parenteral or enteral medication would indicate that intrathecal pump implantation would be unsuccessful. For example, if a patient presents after several months of opioid consumption and has not received significant analgesia or VAS scales of 5 or below, the prescriber must understand that the likelihood of obtaining pain relief from an intrathecal opioid is low. Additionally, if a patient is making frequent emergency room visitations for medication refills or is “prescription shopping”, concerns for addiction and inappropriate use of medications should be raised.

Lastly, compliance is important to have with intrathecal pump therapy. Pump implantation, management and maintenance requires strict compliance to help ensure safe and proper use of the pump. There are serious complications that can happen with routine use of the pump. A motivated patient and good support structure help with the success of the intrathecal pump therapy.

Contraindications to IDDS therapy include patient refusal, noncompliance with visitations and refills, patient anticoagulation, hemodynamic instability, sepsis, and abnormal spinal pathology that may lead to increased intracranial pressure or damage to spinal nerves. Relative contraindications include hypotension, untreated psychiatric conditions, chronic lower extremity edema, and emaciation which would interfere with safe implantation of the device [35].

Comorbidities should be controlled prior to implantation to prevent complications around the time of implantation. Some conditions include but are not limited to bleeding disorders, diabetes mellitus, obstructive sleep apnea, history of infection in and around the area of implantation, MRSA, and immunocompromisation. Immunocompromised patients may require additional antibiotic prophylaxis (i.e., vancomycin).

The trial period is considered to be our best predictor of success with implantable IDDS. Once a patient has been selected to be an intrathecal drug candidate, the health care provider has several questions to answer that will tailor the therapy to the patient. What kind of pain does the patient have? What medication will provide analgesia? How long should the patient be trialed? Currently, there are no universal recommendations or guidelines on how to proceed. The Polyanalgesic Consensus Committee (PACC) of 2012 and the British Pain Society (BPS) of 2008 have expert panel recommendations on how to proceed with the trial. Recommendations by the PACC and BPS are for providers to develop an algorithm that takes into consideration the method of trialing, the trial drug or drugs, duration of trial, patient safety, adequate monitoring, and goals or criteria for implantation [26, 36].

Trials can be done as single/multiple intrathecal injections and/or epidural/intrathecal infusions. The goals are to observe the analgesic efficacy of the drug and to look for any intolerable side effects. Unfortunately, there is no uniform way of performing a trial for intrathecal drug therapy. The Polyanalgesic Consensus Committee of 2012 recommends that the trialing method be based on the patient symptoms (type of pain) and comorbidities [26]. For example, a cancer patient with nociceptive pain that is effectively treated with a drug administered via a more traditional route can likely be done with a single-shot intrathecal dose or no dose at all if the delay in implanting an IDDS will be significant and may even harm the patient. On the other hand, patients that may not clearly benefit from an implantable IDDS can be considered for longer trials and infusions via percutaneous catheters lasting days to weeks to better assess the clinical benefit of implantable IDDS. Therefore, the duration of the trial should be based on a patient to patient basis. Although trials can be done with epidural injections or infusions, the intrathecal route is preferred by most physicians since it will be more correlative to the patient’s physiologic response to the drug.

Currently, the three drugs that are approved by the Food and Drug administration (FDA) to be administered as outpatient intrathecal drugs are baclofen (Gablofen), a GABA-B agonist; morphine (Infumorph), a Phenanthrene opioid agonist; and ziconotide (Prialt), a N-type calcium channel blocker [26]. These drugs can be used individually or in combination of each other depending on the goals of therapy. Additionally, there are off-market or off-label uses of other drugs such as local anesthetics (bupivacaine, ropivacaine, levobupivacaine), NMDA antagonists (Ketamine), alpha-2 receptor agonists (Clonidine), and off-label intrathecal opioids (Fentanyl, Sufentanil, Hydromorphone) [26]. Other drugs such as mixed alpha/beta agonists (epinephrine) and acetylcholinesterase inhibitors (neostigmine) can rarely be used as intrathecal analgesics [8, 9].

In order to pick the best drugs that will help the patient, one must first categorize the type of pain the patient is having. Earlier in the chapter, we classified patients into two categories: Chronic cancer pain and chronic nonmalignant pain patients [7]. This is useful in discussing patient selection for intrathecal pumps but is nonspecific when it comes to choosing the drug therapy. For the purposes of selecting drug therapy, we can classify pain into four categories: nociceptive, neuropathic, mixed (nociceptive + neuropathic) and spastic pain [3].

The most predominant type of pain is nociceptive pain. Nociceptive pain is a noxious stimuli that elicits sharp, shooting, generally well-localized pain and is a result of tissue damage [6, 9]. Generally, most patients with nociceptive will respond to a single-shot test dose of the drug therapy. The PACC of 2012 first-line agents include morphine, ziconotide, hydromorphone, and fentanyl. The choice of drug will be based on the patient’s history of medication use, allergies and intolerances [26].

Neuropathic pain is a noxious stimulus that patients will describe as a burning, lancinating, heavy sensation and/or numbness that is due to nerve damage [3]. A neuropathic pain picture may require a longer trial period to elicit a good analgesic response. This is where percutaneous intrathecal/epidural catheter trials can be useful. These trials could be done as a single/multiple drug bolus trial or an infusion trial lasting a few days to weeks. First-line agents for this setting include morphine alone or with a local anesthetic, most commonly bupivacaine, and ziconotide alone [26]. The decision to use a longer trial will also depend on the patient’s preference and if the patient can tolerate a prolonged infusion.

A mixed pain picture is more challenging to treat. It is important to uncover the symptomatology of the patient and delineate goals of therapy with the patient. This will better prepare the patient for what to expect and which symptoms can be relieved with which medications. Patients with mixed pain are usually treated well with a multimodal analgesic approach. This can include opioids, nonopioids, and local anesthetics. It is likely that oral adjuvants will also be helpful.

Refractory spastic pain from spinal cord injury, cerebral palsy, multiple sclerosis, and amyotrophic lateral sclerosis is treated well with subarachnoid baclofen. Baclofen therapy is initially done as a single-shot test dose. Initial bolus doses range from 25 to 50 μg. If the initial bolus does not elicit significant relief from spasticity, a second dose of 75–100 μg can be administered 24 h after the initial dose. A maximum dose of 100 μg can be given after another 24 h if needed [4, 26, 37].