Neuraxial Analgesia

The ability to administer neuraxial medications has revolutionized oncology pain management in appropriately selected patients.¹, ², ³ In particular, patients with pain refractory to systemic medication who were previously treated with ablative or neurodestructive procedures are now managed with neuraxial analgesia. As a result, the use of neurolytic and ablative procedures in cancer pain management has significantly declined. Neuraxial analgesia will often allow patients to improve their quality of life and to reduce dosage in their current drug regimens⁴ or to derive greater pain relief from their equivalent doses.⁵ The risks of neuraxial analgesia are not zero, but they are less than those accompanying ablative and neurodestructive interventions.

Neuraxial analgesia means drug administration within the spinal canal. The modality involves the injection or infusion of medications into the intrathecal space (also known as the subarachnoid space) or epidural space (Fig. 22.1). The epidural space is located within the spinal canal between the dura and connective tissues covering the vertebrae and ligamentum flavum. Most of the epidural space is filled with fat and does not contain free fluid, unlike the intrathecal space, which contains cerebrospinal fluid (CSF), spinal cord, and nerve roots. The epidural space also contains mixed nerves, because they travel to and from the spinal cord to innervated structures. Medications injected into the epidural space are effective locally on mixed nerves (local anesthetics), may traverse intrathecally and be effective on the spinal cord (primarily hydrophilic opioids such as morphine), or are absorbed systemically in the epidural vascular system and have distant effects (lipophilic opioids such as fentanyl). The intrathecal space, by contrast, is fluid-filled and refers to the area between the arachnoid and pia mater of the brain and spinal cord that contains CSF. The spinal cord is surrounded by three meninges: the dura mater, arachnoid, and pia (Fig. 22.2). The arachnoid and pia, also called the leptomeninges, are separated by the intrathecal space, which contains CSF, arteries and veins, and nerve roots. CSF has an important role in the distribution and elimination of intrathecal medications. It is formed mainly in the choroid plexus of the brain, circulates in the subarachnoid space, and is resorbed into the venous blood through arachnoid villi and through the ependymal lining of the ventricles. In humans, the volume of CSF is replaced three or four times each day at a rate of 0.3-0.4 mL/min. Clinically, the intrathecal space is entered after puncture of the dura mater and free flow of CSF is obtained. Various medications can be injected intrathecally (Table 22.1). These medications act on nerve roots/rootlets (local anesthetics), or bind to synaptic and axonal neuroreceptors in the dorsal horn of the spinal cord, such as µ-receptors (opioids),
calcium channels (calcium channel blockers), or on gamma-aminobutyric acid (GABA) receptors (baclofen, midazolam), where they modulate nociceptive processes.

FIGURE 22.1 Anatomy of the epidural and intrathecal space. (From Morgan, G, Mikhail M and Murray M. Clinical Anesthesiology {ital]. 4th ed., p. 292. New York: McGraw Hill Professional. Copyright Mc-Graw Hill Companies, Inc., 2006.)

FIGURE 22.2 Location of spinal meninges, epidural space, and intrathecal space. (Reused with permission from Marie Dauenheimer, CMI; Ghafoor VL, et al. Intrathecal drug therapy for long-term pain management. Am J Health Syst Pharm. 2007;64:2447-2461.)

The clinical effects of neuraxial medications are influenced usually by the position of the catheter tip⁶ and by the space in which the catheter is placed (epidural vs. intrathecal). Different systems are available to deliver neuraxial medications by continuous infusion. These systems may involve externalized spinal catheters, subcutaneous reservoirs, or implanted infusion pumps (constant rate or programmable).⁷ Although such delivery systems have come into widespread use, particularly for the management of chronic nonmalignant pain, the overall use of long-term intrathecal and epidural opioid administration has been less for cancer pain.

TABLE 22.1 MEDICATIONS USED FOR NEURAXIAL (EPIDURAL, INTRATHECAL) INFUSION

Medication Class	Epidural	Intrathecal
Local anesthetic	Bupivacaine	Bupivacaine
	Ropivacaine
Opioids	Morphine	Morphine
	Hydromorphone	Hydromorphone
	Fentanyl	Sufentanil
	Meperidine	Meperidine
α-2 agonists	Clonidine	Clonidine
N-type calcium channel blocker		Ziconotide
Antispasmodics		Baclofen
Gamma-aminobutyric acid agonist		Midazolam

CHOICE OF SPINAL DELIVERY SYSTEM

A variety of neuraxial delivery systems are available.⁷ The patient’s diagnosis, pain pattern, life expectancy, and the cost of the pain therapy modality have traditionally influenced the selection of a specific drug delivery system. Ever since Behar et al.⁸ first reported on the use of epidural morphine for the treatment of cancer pain, pain management specialists have been looking for a means to overcome the technical problems that can arise during long-term treatment with neuraxial medications. Prominent among these issues is the development of a robust and reliable system of drug administration that is relatively free from complications. Some of the potential complications include leakage and occlusion, catheter dislodgment, infection, and pain on injection or infusion. Attempts to reduce these problems have included tunneling of catheter to the flank or the anterior abdominal wall or implantation of a subcutaneous injection port located on the patient’s anterior chest wall.

Infusion systems may be considered for short-term or long-term administration. The systems commonly used for short-term (usually less than 1 month) are percutaneously inserted into the epidural or intrathecal space without tunneling. Long-term systems are either subcutaneously tunneled and externalized or fully implanted. Fully-implanted systems may have a subcutaneous injection port, an implanted reservoir for continuous infusion at a fixed rate, or a programmable infusion system utilizing a battery powered motor-driven pump. New delivery system technology is in development that will obviate the need for a motor and its high power consumption, thereby reducing replacement costs and surgical risks.

The epidural route for treatment of cancer pain has been associated with a variety of technical complications.⁹ The use of simple percutaneous temporary catheters has been limited. Advantages of percutaneous catheters include ease of placement and ability to attach to an external pump. Disadvantages include the risk of infection with long-term use, lack of dose stability, a high incidence of displacement of the catheter, and fibrosis around the catheter tip resulting in pain on injection and loss of efficacy. A two-piece silicone-rubber exteriorized epidural catheter system described by DuPen et al.¹⁰ allows for specific spinal-level location of the catheter tip from a lumbar entry site and does help reduce the incidence of displacement and pain on injection. Tunneled epidural catheters with subcutaneous injection ports have reduced the complication rate accompanying simple percutaneous and tunneled epidural catheters, particularly for catheter dislodgment and early infections.⁹ In a series of 313 patients, it was noted that the most frequent complications of subcutaneous injection ports were pain on injection (12.0% incidence), occlusion of the portal system (10.9%), infection (8.1%) and leakage of administered medication such that it did not all reach the epidural space (2.1%).11 Possible disadvantages of these systems include the need for skin puncture with each delivery system use and the increased cost of the system.

The intrathecal route is more efficient and potentially less expensive than epidural delivery for the treatment of refractory cancer pain.¹² Advantages of intrathecal administration include better distribution of medication to the target site and enhanced pain relief.¹³ Concerns about externalized intrathecal catheters focus around potentially higher risks of complications (particularly meningitis, CSF fistula, and CSF hygroma) compared to epidural catheters. Clinicians who use the intrathecal route have tended to avoid percutaneously inserted, subcutaneously tunneled, “open” or externalized catheters, preferring a totally implanted system that might reduce the likelihood of these complications. However, such concerns appear to be largely unfounded,¹⁴ particularly with appropriate care of the system.

Fully implanted systems, especially if programmable, are the most convenient for long-term treatment with intrathecal opioids. At this time, only Medtronic has developed a Food and Drug Administration (FDA)-approved programmable pump. However, concomitant use of other medications with opioids significantly complicates the use of these systems particularly when it comes to medication formulation (often requires the use of a compounding pharmacist) and dose adjustment (complex regimens and protocols are required). The issue of dose adjustments and associated complexities is particularly challenging in the changing environment of the cancer patient with progressive disease and associated pain.

TABLE 22.2 RECOMMENDED MEDICATION DOSES AND CONCENTRATIONS FOR INTRATHECAL THERAPY

Drug	Standard Concentration	Maximum Recommended Compounded Concentration	Typical Starting Daily Dose	Maximum Dose/Day
Baclofen	50 µg/mL, 0.5 mg/mL, 2 mg/mL	2 mg/mL	100 µg	∼2000 µg
Bupivacaine	0.25%, 0.5%, 0.75%	40 mg/mL	48 mg	300 mg
Clonidine	100 µg/mL, 500 µg/mL	2 mg/mL	50 µg	1.5 mg
Fentanyl	50 µg/mL	6000 µg/mL	∼100 µg	∼1500 µg
Hydromorphone		10 mg/mL	0.25 mg	4 mg
Morphine	10 mg/mL, 25 mg/mL	30 mg/mL	1 mg (opioid tolerant), 0.5 mg (opioid naive)	15 mg
Sufentanil	50 µg/mL	50 µg/mL	1 µg (opioid tolerant), 0.1 µg (opioid naive)	15 µg
Ziconotide	25 µg/mL, 100 µg/mL	2 mg/mL	1.4 µg	19 µg

The use of implantable infusion technology is not without problems; complication rates varying from 5% to 15% have been reported.¹¹,¹⁵, ¹⁶, ¹⁷, ¹⁸ Overall, these systems appear to offer a lower risk of certain complications (infection, displacement, and pain on injection) than other systems but may also introduce specific delivery system problems (battery failure, pump torsion, overfilling of reservoir, programming errors, and so on).¹⁹ Furthermore, the pump cannot be safely implanted in a radiation therapy portal. In our experience, the ideal system in the oncology patient with frequently changing analgesic needs is an externalized, implanted, intrathecal system connected to an external pump and managed at home through a home infusion company. Expensive hardware is avoided; home care has been dependable and safe in our environment.

CHOICE OF NEURAXIAL MEDICATION

The efficacy of neuraxial opioid therapy for both chronic nonmalignant pain and cancer pain is well documented but it has its limitations (vide infra).²⁰ Stearns et al.²¹ proposed an algorithmic approach to intrathecal drug delivery for the management of cancer pain. Medications suggested for use included various opioids, bupivacaine, baclofen, clonidine, droperidol, ketamine, and midazolam. Table 22.2 lists our recommendations for intrathecal drug delivery. Intrathecal morphine has a long safety record and is the only µ agonist approved by the FDA for long-term administration. At this time, no drugs other than morphine, ziconotide, and baclofen are
FDA-approved for intrathecal use. Local anesthetics have been used with increasing frequency for refractory pain control but have limitations, particularly with regards to lower extremity function and bladder/bowel control. Baclofen, although safe and approved for long-term use by the FDA, has little effect in pain management, unless pain is associated with spasticity, and should not be tried for other pain-related complaints. The usefulness of clonidine is unclear, and its effects on the circulatory system may be unpredictable and difficult to manage.²² Ziconotide is FDA-approved for intrathecal use but its side effects have been problematic. Long-term use of unapproved medications intrathecally should be considered experimental as their safety profiles are unproven by this route of administration, in spite of widespread use in nonmalignant pain states. In light of recent associations of intrathecal medications with granuloma formation,²³, ²⁴ patients who require high-dose or high-concentration intrathecal opioid therapy and those who receive drugs or admixtures not approved for intrathecal use should be monitored closely for signs of an extra-axial mass or catheter malfunction.²⁴

NEURAXIAL OPIOID THERAPY

Intractable cancer pain has been routinely treated with the neuraxial (intrathecal or epidural) infusion of opioids for more than a quarter-century. Spinal and supraspinal sites are the areas at which intrathecal or epidural morphine injections are effective.²⁵ Initial reports indicated high efficacy and few complications or side effects.²⁶, ²⁷, ²⁸, ²⁹, ³⁰ At that time, many believed that the solution to intractable pain was in hand and widespread application awaited only the development of suitable long-term delivery systems.³¹ Subsequently, additional case series and surveys made it clear that neuraxial administration of opioids was not risk free,³² that tolerance to neuraxial opioids could be as big a problem as tolerance to oral or systemic opioids,³³, ³⁴, ³⁵ and that not all patients with pain from cancer experienced good pain relief with neuraxial opioids.¹⁸,³⁶, ³⁷ The failure rate of epidural and intrathecal opioids in cancer pain can be as high as 30%.³⁸, ³⁹ Techniques of drug delivery also seemed to play a role in outcomes. Gourlay et al⁴⁰ reported that bolus dosing of neuraxial opioids resulted in less tolerance than a continuous infusion. On the other hand, Hassenbusch et al.⁴¹ reported that chronic epidural morphine infusions were associated with very little tolerance development. As the field of neuraxial opioids developed, it became clearer that there were distinct differences between epidural and intrathecal administration, some related to drug pharmacokinetics, others to anatomy and technique advances and that the intrathecal route was better suited to treatment at home by continuous infusion than the epidural route.⁴² Route of administration is now recognized as a critical factor for neuraxial opioid administration.

Epidural Administration

Although the objective of epidural administration is to affect pain relief by altering spinal cord dorsal horn synaptic transmission, there is some systemic absorption by all epidurally-administered drugs, including opioids.⁴³ One of the most important factors governing the spinal bioavailability of epidurally-administered opioids is their ability to redistribute out of the epidural space and into surrounding tissues. The kinetics of drug movements among epidural, intrathecal, and plasma compartments are complex, and the epidural and intrathecal spaces cannot be viewed simply as two compartments separated by a single barrier.⁴⁴ CSF and plasma pharmacokinetics of opioids do not parallel their epidural pharmacokinetics and the hydrophobic character of opioids governs multiple aspects of their lumbar epidural pharmacokinetics. A significantly larger proportion of the epidural dose of morphine reaches the CSF than is the case for any of the other more lipid-soluble opioids.⁴⁴

The utility of each drug to be administered epidurally must be determined by studying whether the observed analgesia is due to selective action on the spinal dorsal horn or is related to systemic uptake of the drug and effects upon sites other than the spinal cord, such as the brain or peripheral nerves. Obviously, combinations of these two effects are likely for many opioids. Morphine is the standard drug against which all others are compared for efficacy and side effects. It was the first opioid administered into the epidural space that was active on the dorsal horn. Studies have shown conclusively that it produces analgesia by its actions on the dorsal horn of the spinal cord.⁴⁵ Subsequent studies have shown that highly lipophilic opioids, such as fentanyl, sufentanil, or alfentanil are inappropriate for epidural administration because of their rapid systemic uptake and effects outside of the spinal dorsal horn.⁴⁶, ⁴⁷, ⁴⁸, ⁴⁹, ⁵⁰

Reports documenting the use of different opioids in the epidural space are now available; morphine is clearly the drug of choice for the majority of patients for this application. However, the long-term superiority to systemically administered drugs has been debated. In 1989, Sjogren reported no significant differences between oral and epidural routes of administration in long-term treatment of 14 cancer patients when pain relief, sedation, and reaction times were assessed.⁵¹ Hassenbusch et al. studied the efficacy and safety of continuous epidural morphine administration in 69 patients with cancer pain in the midline abdomen/pelvis or lower extremities who did not respond adequately to systemic opioid treatment.⁵² The detailed characteristics of the pain syndromes were not described for each patient. An epidural catheter was placed for evaluation of efficacy of morphine: 60% of this group reported good pain relief and had an internalized catheter and Infusaid pump (a pump with a fixed infusion rate) implanted. Before the administration of epidural drug, the patients reported a visual analog scale (VAS) of 8.6 ± 0.3. One month after implantation of the catheter and pump, the VAS scores were 3.8 ± 0.4 (P < .001). Long-term benefits were described: 80% of the implanted patients reported satisfactory pain relief (at least a 30% reduction from preimplant VAS levels) at 3 and 6 month reevaluation. Also, the systemic opioid requirements were significantly decreased compared to preimplantation levels: by 79% at 1 month and 64% at 9 months. Some tolerance was seen: the mean epidural dose was 20.7 ± 2.6 mg/day 1 month after implantation and rose to 49.3 ± 9.9 mg/day 9 months after initiation of this treatment strategy. This study had a very low complication rate and reported no
instances of respiratory depression, epidural infection, meningitis, epidural scarring, or catheter failure. This was one of the seminal studies that demonstrated good longterm results with epidural morphine infusion in patients with pain from cancer.

By contrast, other studies have revealed a much higher failure rate with morphine.³⁸, ³⁹ Nine years’ experience with epidural morphine in 146 cancer pain patients was reported by Samuelsson et al.⁵³ The mean treatment time for each patient was 92 days and epidural opioids were the sole pain treatment in 53% of patients. They reported that the mean dose of epidural morphine was 18 mg/day with a range of 6 to 120 mg/day. At the termination of this treatment strategy, the mean daily dose was 69 mg/day with a range of 2 to 540 mg/day. Absence of an initial satisfactory response to epidural morphine was noted in 25 patients. Withdrawal from initially successful therapy was observed in 27 patients; an additional nine patients experienced catheter failures and five reported drugrelated reasons for termination of epidural morphine. Those most likely not to respond to epidural morphine were patients with neuropathic pain states, those with movement-induced incident pain, those with visceral pain and those with painful cutaneous ulcerations.

Our review of the literature and our own experiences suggest that epidural opioid administration is not likely to be a successful long-term therapy for pain associated with cancer, particularly if it is the sole analgesic therapy. The major issue is the development of fibrosis around the epidural catheter that impedes drug delivery and results in local pain at the time of injection.⁵⁴ Tolerance does occur, but it is not usually an unmanageable problem. Maintenance of the catheter in a favorable position can also be a problem. We strongly prefer intrathecal administration of drugs in setting of cancer pain, as we believe it a more reliable method of drug delivery and it does not, in our hands, have increased risks.

Intrathecal Administration

Intrathecal drug delivery offers significant advantages for pain relief over epidural administration of opioids. These include: uniform drug distribution within the CSF (as long as highly lipophilic drugs are not used), superior efficacy of pain relief, relative absence of systemic effects at clinically useful intrathecal doses, ease and certainty of catheter insertion by CSF aspiration or contrast agent injection, and lower reported incidence of complications. As in the epidural space, the behavior of opioids in the intrathecal space is governed largely by their lipid solubility. Hydrophobic opioids have a very large apparent volume of distribution compared with that of more hydrophilic opioids.⁵⁵ Rapid elimination of hydrophobic opioids from the CSF results in very limited rostral spread in CSF. In addition, the bioavailability of these opioids at spinal cord sites more rostral than the site of administration is also very limited. This explains why delayed respiratory depression is observed clinically with morphine but not with fentanyl or sufentanil.

The ratio of intrathecal dosing to oral dosing is approximately 1 to 300, thereby significantly reducing the systemic opioid exposure. Implanted catheters and pumps can be used to infuse any opioid in aqueous solution, including morphine, hydromorphone, methadone, meperidine, fentanyl, and sufentanil. The intrathecal delivery route has been the subject of numerous reports with somewhat differing conclusions.²,¹⁵,¹⁸,³⁶,⁵²,⁵⁶, ⁵⁷, ⁵⁸, ⁵⁹, ⁶⁰, ⁶¹, ⁶², ⁶³, ⁶⁴ Most of these reports describe significant pain relief from intrathecal opioids, but this literature suffers from many deficiencies and a complete picture of the use of this technology is often difficult to ascertain. Few studies have defined pain relief in a prospective fashion using standardized instruments. Some studies have described a few patients who manifested no pain relief to intrathecal opioids in spite of often astronomical increments in dosing.60,63,65-66 In most cases we do not know if this was a failure of the delivery system or a pain that did not respond to opioid treatment. The actual rate of patients with cancer pain not responsive to intrathecal opioids is not known, but it is certainly not zero. Nor do we know with any precision the efficacy of intrathecal opioids for the relief of cancer pain.

A useful study by Paice⁶⁰ contrasted the pain relief from intrathecal opioids in patients with cancer and nonmalignant diseases. The best results were seen in patients with cancer pain of somatic origin. However, when compared to non-cancer pain patients, the average initial dose of opioid was higher in the patients with cancer. Whereas the average dose used by cancer patients rapidly increased and then reached a plateau, non-cancer patients seemed to exhibit a more gradual but constant dose increase over long periods of use. This study contrasted with that of Yaksh and Onofrio⁶⁴ who studied the changes in morphine infusion dosing in 130 cancer pain patients and reported that the initial average dose of 4.8 ±0.4 mg/day increased to 21 ± 9 mg/day after 1 year of therapy. Plummer’s study in 1991¹¹ indicated wide variations in the dose requirements to adequately control cancer pain. They indicated that the type of pain also played a significant role in the response to intrathecal opioids, similar to what was reported above for epidural opioids. Long-term intrathecal morphine infusions were studied by Gestin et al.⁶⁷ in 50 patients with advanced cancer and pain. The average infusion lasted 142 (range: 7 to 584) days and the mean starting dose was 2.5 (range: 0.4 to 8.3) mg/day; this increased to a mean final dose of 9.2 (range: 1 to 94) mg/day. The average dose was 5.4 (range: 1 to 23) mg/day and all the patients were said to have satisfactory pain relief. One interesting report suggested that methadone given intrathecally produced pain relief inferior to that seen with morphine, even with doses ten times the equivalent of morphine.⁶⁸

Adverse Effects of Neuraxial Opioid Therapy

The adverse effects of systemic morphine are well described in the medical literature. Neuraxial administration of this drug results in a similar range of side effects, despite the more limited distribution of the drug in the body. Some of these side effects are temporary, normally lasting for the first several days after initiation of therapy and then resolving; others are more enduring.⁶⁹ In a retrospective study of 82 patients receiving long-term opioid therapy for noncancer-pain, Winkelmuller and Winkelmuller⁷⁰ reported the most frequent side effects as
constipation (50%), disturbance of micturition (42.7%), and nausea (36.6%). These occurred early in the course of therapy and responded to appropriate medication. Some patients experienced a loss of libido or amenorrhea for the first 6 to 8 months of therapy, but these side effects proved self-limiting to most patients and disappeared after 12 to 14 months of therapy. In most series, common problems associated with intrathecal opioid therapy have included nausea, vomiting, pruritus, urinary retention, and constipation. Less common adverse events have been respiratory depression and opioid-induced hyperalgesia. Less frequently recognized by clinicians are problems associated with endocrine abnormalities (in particular, sexual dysfunction) and edema.

Abs et al.⁷¹ reported on the endocrine consequences of long-term intrathecal opioid use and noted that, in patients receiving intrathecal opioids, the majority of men and all women developed hypogonadotropic hypogonadism, about 15% developed central hypocorticism, and about 15% developed growth hormone (GH) deficiency. In this study of 73 patients (29 men and 44 women; mean age, 49.2 ± 11.7 year) receiving intrathecal opioids, the mean duration of opioid treatment was 26.6±16.3 months and the mean daily dose of morphine was 4.8 (3.2 mg). Decreased libido or impotency was present in 23 of 24 men receiving opioids. Serum testosterone level was less than 9 nmol/L in 25 of 29 men and was significantly lower than that in the control group (P < .001). The free androgen index was below normal in 18 of 29 men and was significantly lower than that in the control group (P < .001). The serum luteinizing hormone (LH) level was less than 2 U/L in 20 of 29 men and was significantly lower than that in the control group (P < .001). Serum folliclestimulating hormone (FSH) was comparable in both groups. Decreased libido was present in 22 of 32 women receiving opioids. All 21 premenopausal women developed either amenorrhea or an irregular menstrual cycle, with ovulation in only one. Serum LH, estradiol, and progesterone levels were lower in the opioid group. In all 18 postmenopausal women, significantly decreased serum LH (P < .001) and FSH (P = .012) levels were found. The 24-hour-urinary free cortisol excretion was below 20 microg/day in 14 of 71 opioid patients and was significantly lower than that in the control group (P = .003). The peak cortisol response to insulin-induced hypoglycemia was below 180 µg/L in 9 of 61 opioid patients and was significantly lower than that in the nonopioid group (P = .002). The insulin-like growth factor I standard deviation (SD) score was below -2 SD in 12 of 73 opioid patients and was significantly lower than that in the control group (P = .002). The peak growth GH response to hypoglycemia was below 3 µg/L in 9 of 62 subjects and was significantly lower than that in the control group (P = .010). Thyroid function tests were normal. No other metabolic disturbances were noted, apart from significantly decreased high-density lipoprotein cholesterol levels (P = .041) and elevated total/high density lipoprotein cholesterol ratio (P = .008) in the opioid group compared to the control group. Supplementation with gonadal steroids improved sexual function in most patients.

Water retention and peripheral edema caused by intrathecal opioids can be problematic, resulting in limitations of physical activity and the production of lymphedema, ulcerations, and hyperpigmentation of the skin.⁷² Cephalad migration of opioids within the CSF and interaction with opioid receptors in the posterior pituitary gland with the release of vasopressin may explain opioidinduced fluid retention.⁶⁹ We have had patients whose fluid retention was so profound that their intrathecal opioids had to be discontinued.

Catheter-tip inflammatory-like masses or granulomas have been attributed to implanted intrathecal catheters used for long-term pain management (Fig. 22.3).²⁴ The first case was reported by North et al. in 1991.⁷³ Drug that precipitates from solution can cause an inflammatory mass and mimic a tumor in the subarachnoid space.⁷⁴ Preliminary experimental evidence suggests that dural mast cell activation and subsequent release of inflammatory mediators (such as cytokines and histamine) are responsible for catheter tip inflammatory mass lesions.²³,⁷⁵ Cases have occurred in patients receiving commercially prepared preservative-free morphine sulfate alone, as well as formulations of hydromorphone and morphine sulfate alone or in combinations with other analgesics and local anesthetics prepared by compounding pharmacies. The incidence of granuloma formation has been estimated to be 0.04% after 1 year of therapy, increasing to 1.15% after 6 years of therapy.⁷⁶ We suspect that the actual rate of granuloma formation is significantly higher; many cases have probably not been reported to central data banks and smaller granuloma may not be clinically apparent. High CSF drug levels of opioids, and in particular, drug concentration infused rather than absolute dose have been implicated as the cause of the inflammatory masses.⁷⁷ Yaksh et al.⁷⁸ noted 30% of granulomas developed in patients receiving less than 10 mg/day; 40% were using morphine concentrations of less than 25 mg/mL. Many of these patients were receiving drug mixtures, which may also be a key factor in granuloma formation.

Symptoms and signs consistent with inflammatory masses include diminished or loss of drug (analgesic) effect, new-onset radicular and back pain, and associated spinal cord neurologic deficits. Overt findings may include changes in the patient’s baseline neurologic condition, such as motor weakness (including gait difficulties), sensory loss (including proprioceptive loss), hyper- or hypoactive lower extremity reflexes, and any evidence of bowel or bladder sphincter dysfunction.⁷⁹ Imaging with a gadolinium-enhanced magnetic-resonance imaging (MRI) or computed tomography (CT) myelogram confirms the diagnosis. Plain MRI or CT scan or x-ray is of no value in establishing this diagnosis. Clinical evaluation based solely on the development of neurologic symptoms or decreasing analgesic efficacy alone in the absence of specific imaging studies lacks sufficient sensitivity to exclude the presence of catheter-associated masses.⁸⁰ Subtle prodromal symptoms and signs may include diminishing analgesic effects (loss of previously satisfactory pain relief) and remarkable or unusual increases in the patient’s underlying pain. Some patients may manifest frequent dose escalations in the attempt to recapture analgesic effects. In other cases, dose escalations and sizable drug boluses reduce the patient’s pain only temporarily or to a lesser degree than previous experience predicted. Gradual, insidious neurologic deterioration weeks or months after the
appearance of subjective symptoms was the most common clinical course before the onset of myelopathy or cauda equina syndrome.²⁴ Catheter tip granuloma formation is a potential neurologic disaster and must be considered as an emergent problem in any patient who manifests new symptoms and signs of pain or spinal cord dysfunction.

FIGURE 22.3 A 51-year-old woman who received an intrathecal catheter (tip T8/9) for failed back syndrome. Four years later, patient presented with severe radicular back pain at T8 level. Sagittal MRI (T1, fat suppressed, post gadolinium) shows a granuloma at T8 level (A)(arrow). Axial MRI (B) shows the granuloma dorsal to the cord (arrow). The catheter was surgically removed with gradual resolution of symptoms. MRI, Magnetic resonance imaging.

Escalating doses of intrathecal morphine may prove problematic because of increased risk for hyperalgesia and myoclonus.⁵⁷ Opioid-induced hyperalgesia is a pronociceptive phenomenon that can occur with acute or chronic opioid administration. This paradoxical effect is likely due to the opioid effects on a variety of cellular targets including N-methyl-D-aspartic acid (NMDA) receptors, spinal glutamate transporters, and protein kinase C81-82 and does not appear to be an opioid-receptor mediated event.⁸³ Ibuki et al.⁸⁴ demonstrated the increased content of excitatory amino acid neurotransmitter in spinal cord tissue with intrathecal morphine administration. Sakurada et al.⁸³ suggests that nociception associated with high-dose intrathecal morphine may mediated by morphine-3-glucuronide. Opioid-induced hyperalgesia is discussed extensively in Chapter 19.

Bolus Versus Continuous Infusion of Opioids

The optimal method of delivery of neuraxial opioids (bolus vs. continuous infusion) is unclear. Controversies exist regarding the development of tolerance, differences in quality of pain relief, and duration of effect between the two modalities of administration. The behavior of opioids when administered as a slow continuous infusion is different from single injection bolus.⁴³ As a slow infusion, the concentration of morphine is limited to a short distance from the point of administration. This implies that the CSF is a very poorly mixed compartment. By contrast, a single bolus injection of intrathecal morphine has a prolonged and extensive neuraxial effect.⁸⁵

Clinical effectiveness of intermittent bolus and continuous infusion of epidural opioids may not differ. In a study of 28 patients with severe cancer pain, these modalities were compared.⁴⁰ The mean duration of treatment was 169 (range: 6 to 537) days for continuous infusion patients and 140 (range: 28 to 378) days for repeated bolus doses. Both treatment modalities provided adequate pain control and the modalities were judged equally effective. Rauck et al.⁸⁶ evaluated the efficacy and safety of an intrathecal bolus delivery system of morphine in 119 cancer patients with cancer pain refractory to systemic opioids or uncontrollable opioid-related side effects. Pain relief, reduction in systemic opioid use, and reduction in opioid-related complications were evaluated together as a measure of overall success over a 16-month period. Patients were eligible if they had pain from cancer or cancer therapies or experienced pain that could only be controlled at doses that caused intolerable side effects. In addition, patients were enrolled if they had a life expectancy of 4 months or greater. Although not reported directly, it appears that there were only nine patients remaining in the study by month 16. The median starting bolus was 1 mg of morphine and the median average daily dose was 1.8 mg/day.
Median bolus size and median average daily dose increased steadily to 2.5 mg and 5.1 mg/day, respectively, at 4 months. Average numeric analog scale pain decreased from 6.1 to 4.2 at 1 month and was maintained through month 7 (P < .01) and through month 13 (P < .05). Systemic opioid use was significantly decreased throughout the study (P < .01). The majority of patients in this present study decreased their overall opioid requirements by greater than 50% when using intrathecal opioids; many discontinued systemic use entirely. There was a pronounced decrease in the incidence of opioid-related side effects with intermittent intrathecal dosing compared to previous systemic dosing. Overall success (>50% reduction in numeric VAS pain score, use of systemic opioids, or opioid complication severity index) was reported in 83%, 90%, 85%, and 91% of patients at months 1, 2, 3, and 4, respectively. The most commonly recorded complications were sleep disorders, daytime drowsiness, constipation, and nausea. Frequency and severity of sleep disorder, daytime drowsiness, urinary retention, and constipation were significantly improved by 1 month, and five of the eight complications were significantly improved at 4 months.

Tolerance to the antinociceptive effect of morphine is well established in most mammals.⁸⁷ Tolerance occurs for a number of reasons, including previous opioid exposure, changes in pain modality or intensity, and psychological factors. However, in patients with cancer, it may be difficult to determine the precise reason for dose increase because of the often progressive nature of the disease and associated increase in pain. Systemic morphine via repeated subcutaneous injections can result in substantial tolerance within 3 days.⁸⁷ By contrast, tolerance develops more slowly to morphine at localized spinal sites.35,88-89 Animal studies with chronic intrathecal catheters and continuous infusion of morphine over a period of days demonstrate a dose-dependent antinociception that declines over the infusion interval implying the development of opioid tolerance.⁹⁰ Tolerance produced by morphine infusion is dependent on an increase in local phosphorylating activity by protein kinase C in the dorsal horn of the spinal cord.⁹¹ Human studies of repeated intrathecal or epidural morphine administration have produced conflicting results. Ventafridda et al.¹⁸ observed a rapid loss of effectiveness of 1 mg of morphine during a 3- to 5-day sequence of once-daily intrathecal injections in cancer patients. In a retrospective study of the doses of morphine administered intrathecally by chronic infusion in a population of 163 patients (130 patients had pain that was metastatic in origin) treated by 19 physicians, Yaksh and Onofrio⁶⁴ reported a time-dependent increase from 4.8 ± 0.4 mg/day at week 1 (n = 130) to 16 ± 4 mg at 24 weeks (n = 33) and to 21 ± 9 mg at 52 weeks (n = 10). However, 48% of patients infused for periods in excess of 3 months showed less than a twofold increase in dose by 3 months. In a series of 35 patients treated with continuous intrathecal morphine (mean = 5.4 months) for intractable cancer pain, Penn and Paice⁶¹ reported that the occurrences of tolerance were infrequent and could be managed effectively. A retrospective review of 159 patients with refractory cancer pain treated with daily boluses of intrathecal morphine demonstrated only a moderate increase in intrathecal opioid dose that did not limit the patients’ ability to obtain adequate relief.⁹² The mean follow-up period was 95 days (range: 5 to 909 days), the mean starting dose of intrathecal morphine was 2.69 mg (range: 1 to 7.5 mg), and the mean terminal dose was 7.82 mg (range: 1 to 80 mg).

In summary, these studies suggest that, although tolerance to the effect of spinal morphine may develop in the cancer patient, tumor progression is much more likely to increase opioid requirements in a patient with opioidresponsive pain.

Assessment of Long-Term Efficacy of Neuraxial Opioids

The long-term benefits of epidural and intrathecal opioid therapy for cancer pain have been difficult to objectively assess for many reasons:

The criteria for treatment success are often vague and frequently vary between reports. There have been no standardized outcomes criteria.
The patient selection criteria for initiation of neuraxial opioid therapy have varied widely. In some centers, neurologists or oncologists are primarily responsible for pain management. They may be more content with systemic analgesics and more conservative with invasive measures than pain management specialists. They may refer patients for consideration of neuraxial therapies late in the course of the illness, when metastases have already become widespread, nociceptive inputs are high, and tolerance can be a significant problem. Frequently, such patients require supplementary local anesthetics for pain control; opioids alone are inadequate. By contrast, other centers, where pain management specialists have primary responsibility of cancer pain management, may initiate spinal therapy at an early stage, possibly even when systemic therapy may be equally if not more effective.
Cancer patients frequently have more than one type or location of pain, and the outcomes data are not specifically related to an identified pain site or type.
Meaningful statistical inferences cannot be made from the data as presented.

Some patients who experience inadequate pain relief and/or intolerable side effects from systemic opioids can obtain pain relief without side effects with intrathecal or epidural administration of the same or different drugs. There will be additional patients who never experience adequate pain relief with neuraxial opioids (so called “treatment failures”—vide infra). Different types of pain in patients with cancer may respond differentially to spinal opioids.⁹³ Arner and Arner⁹⁴ studied the differential effects of epidural morphine in patients with cancer pain. They utilized intermittent bolus doses of morphine in 55 patients with different patterns of cancer pain (daily dose range: 4 to 480 mg; dose interval range: 3 to 12 hours). Twenty-eight of the 55 patients (51%) became pain free. Another 21 patients were completely relieved of one of two pain types, whereas other coexisting pain types were either unaffected or only partially relieved. Six patients reported no pain relief. Continuous pain originating from deep somatic structures led to the best outcomes with epidural morphine. Continuous visceral pain or
intermittent somatic pain originating from such things as a pathologic fracture, led to less predictable outcomes. Pains that were cutaneous in origin, thought to be neuropathic, or secondary to intestinal obstruction rarely responded. Coexisting pains that were not related to the underlying malignancy were not alleviated at all in ten patients. This study and others⁵³ suggested that there was some predictability for pain relief based upon the character of the pain after treatment with epidural morphine.

Summary of Neuraxial Opioid Use in Cancer Pain

Current knowledge on the use of spinal opioids in cancerrelated pain allows us to make the following generalizations, based upon the literature and our own experience of cancer patients with pain:

It is common to see a slow increase in mean daily opioid dose requirements in many patients, often at a rate that doubles every few months. Some patients may escalate their dose requirements even more rapidly, particularly in the terminal phase of their disease.⁹⁵
Dose escalation tends to be slower with intrathecal than epidural administration systems. This may relate to the tendency for fibrosis to occur around epidural catheters.¹⁶
The escalation of dose over time tends to be slower in patients receiving continuous infusions as opposed to intermittent bolus injections.⁹⁶ This may be related to the more rapid development of tolerance when spinal opioid receptors are exposed intermittently to high opioid doses.
Higher spinal opioid doses are required for pain that is neuropathic in origin than for somatogenic pain.94 Opioid requirements for visceral pain are usually intermediate between neuropathic and somatic. There is no type of pain that never or always responds to neuraxial opioids.
Incident pain (occurring with activity) is difficult to control with neuraxial opioids.⁸⁶
Higher systemic opioid requirements before implementation of spinal opioids are associated with higher neuraxial requirements.⁹⁷
Many cancer pain patients who have not obtained pain relief with systemic opioids eventually demonstrate little benefit from epidural or intrathecal opioids. Although pain relief is adequate in 45% to 90% of patients with initiation of intraspinal opioid therapy, dose escalation and increased need for supplemental oral opioid therapy occur commonly over the first few months of therapy.¹⁸,³⁶,⁶⁴,⁹⁴,⁹⁶ Van Dongen et al.98 noted that 33% of patients failed intrathecal morphine therapy (after intrathecal therapy started because of failure of epidural therapy) and required the addition of bupivacaine for improved pain control.
Before the widespread use of intrathecal opioids, the most commonly utilized ablative procedure for pain from cancer was cordotomy. This operation was originally done via a laminectomy, but in the late 20th century percutaneous methods for cervical cordotomy were developed. The long-term results for pain relief after cordotomy (there are only case series, not cohort studies or high quality trials) appear to be better than those for neuraxial opioids and the quality of pain relief also seems superior. Of course, the initial complication rate was also higher for cordotomy. However, health care is largely a social convention; today, there are very few cordotomies performed for cancer pain in the United States and neuraxial opioids are the standard when oral or parenteral medications are inadequate.

Intracerebroventricular Opioids

It is possible to deliver opioids directly into the cerebral ventricles using an intracerebroventricular (ICV) catheter attached to a subcutaneous reservoir. A neurosurgeon is required for catheter placement. Morphine sulfate has been the most commonly administered drug; it has a dramatic increase in potency in the ICV route when compared to intrathecal or epidural infusions. The ICV route may affect supraspinal pathways for analgesia that are not accessible to spinally administered opioids.⁹⁹ ICV morphine doses have been reported from 50 to 700 µg/day.¹⁰⁰, ¹⁰¹ The most commonly utilized drug administration system consists of a battery-driven infusion pump, implanted subcutaneously in the anterior abdominal or chest wall, and connected by subcutaneous tubing to the ventricular catheter that delivers the drug into the lateral ventricle.

Another administration strategy has been to connect the ICV catheter to an Ommaya reservoir and utilize intermittent percutaneous drug injections. ICV injections appear to last significantly longer than intraspinally administered drug. Some patients have been reported to obtain excellent pain relief via an implanted ventricular catheter connected to a subcutaneous Ommaya reservoir with 1 to 2 injections per day.¹⁵ ICV drug delivery is particularly useful in head and neck cancer pain. Its use has also been reported in patients who initially manifested a favorable response to intraspinal infusions of opioids and subsequently development tolerance and were nearing the end of life. ICV injections or infusions manifest similar safety and side effects to intraspinal infusions, with the exception that an increased risk of respiratory depression exists during the first 3 days of therapy.¹⁵

Some refractory head and neck cancer pains do not respond to spinal opioids but will respond to ICV opioids. By contrast, pains below the waist are more amenable to spinal drug administration. Ballantyne et al.¹⁰² performed a meta-analysis of 1,587 cancer patients, comparing ICV to the more commonly performed epidural and intrathecal opioid administration. All patients included had intractable cancer pain that had not adequately responded to systemic treatment. This study showed that sedation and confusion was observed in 4% to 5% of patients receiving ICV therapy. Pruritus, urinary retention, and persistent nausea occurred more frequently with spinal administration than with ICV therapy. Patients receiving ICV morphine required an initial dose of 0.25 to 2.0 mg/day. Onset of pain relief occurred in 2 to 30 minutes, and the typical duration of pain relief after a single dose was 12 to 48 hours. Tolerance seemed to be less of a problem than with either epidural or intrathecal therapy. Dose escalation was very gradual (the average increase in daily dosage was 0.375 mg/month). In a subsequent study,
Ballantyne et al.¹⁰³ reported on 72 uncontrolled trials in cancer patients assessing ICV (13 trials, 337 patients), epidural catheters (31 trials, 1343 patients), and subarachnoid catheters (28 trials, 722 patients). These uncontrolled studies suggested excellent pain relief in 73% of ICV patients compared with 72% of those with epidural and 62% of those with subarachnoid catheters. Unsatisfactory pain relief was infrequent in all three treatment groups. Side effects, commonly persistent nausea, persistent and transient urinary retention, transient pruritus, and constipation occurred more frequently in patients with epidural and subarachnoid catheters. On the other hand, respiratory depression, sedation, and confusion were more common with ICV. The incidence of major infection when implanted pumps were used with epidural and subarachnoid catheters was zero. Reported complications with ICV therapy were less than with epidural or subarachnoid catheters. ICV administration of opioids should be considered in patients with head and neck cancer pain that does not respond adequately to intraspinal delivery of drug.

INTRATHECAL THERAPY WITH BUPIVACAINE AND MORPHINE

The use of neuraxial opioids with local anesthetic agents is well established for postoperative pain management.¹⁰⁴, ¹⁰⁵, ¹⁰⁶, ¹⁰⁷, ¹⁰⁸ Bupivacaine-induced adverse effects including sensory deficits, motor complaints (especially lower-extremity impairment), bladder/bowel dysfunction, and signs of autonomic dysfunction led to concerns about these effects that have limited the consideration of this modality for home care in cancer patients. We now know that this form of neuraxial therapy can be safely administered in the outpatient setting. The safety and efficacy of long-term epidural bupivacaine-opioid infusion in 68 patients with cancer pain refractory to epidural opioids alone has been demonstrated.¹⁰⁹ The majority of these patients experienced pain relief with little or no sympathetic or sensorimotor impairment after the first 24 hours with bupivacaine concentrations of varying from 0.125% to 0.25%, and could receive management at home or in chronic care settings without the need for rehospitalization. Transient postural hypotension during the first 24 hours occurred in only 9% of patients and was easily managed with close monitoring and fluid therapy during initiation of therapy.

The number of patients who will require neuraxial therapy with local anesthetics in addition to opioids is unknown. Reasons for selection of patients vary and therapy can be started before systemic opioid therapy has been optimized. The usual indications for intrathecal bupivacaine and morphine therapy are progressive pain not responding to opioid therapy or unacceptable opioidrelated side effects in patients who have satisfactory pain control. Clinicians should also distinguish the need for intrathecal therapy over epidural. Certain tumor-related pain problems predictably are less likely to respond to epidural analgesia.¹³ These include mucocutaneous ulcers, pain caused by body movements as a result of fracture, edematous swelling and/or ischemia of an extremity, and neuropathic pain from plexus infiltration. In these situations, the administration of intrathecal morphine (IT-morphine) alone does not appear to offer any significant advantages.⁹⁵ Different sites of drug action and synergistic effects of bupivacaine and morphine are also important for refractory pain problems.¹¹⁰ IT-bupivacaine with its predilection for unmyelinated axons could contribute to the clinical efficacy by blocking slow, “burning” pains¹¹¹ which may not respond to morphine alone.

In addition, the combination may delay the occurrence of tolerance to IT-morphine because of the smaller doses and the added analgesic effects of bupivacaine. The route of administration is also important and favors the intrathecal over epidural route.¹³ Better and more predictable drug distribution to the required area is probably the primary reason why intrathecal medications are more effective. Distribution of effect after epidural injection of local anesthetics varies widely and may only be partially predicted based upon known factors.¹¹² Epidural anesthesia has an unpredictable extent and duration. Differences in the surface area of the lumbosacral dura, epidural fat volume, and epidural venous plexus velocity might explain the variability in the extent and duration of epidural anesthesia.¹¹³ Cranial spread of lumbar epidural anesthesia is more extensive than caudad spread.¹¹⁴ Block of the fifth lumbar and first sacral segments by anesthetics is often delayed or incomplete¹¹⁵ and may be related to the large size of these nerve roots.¹¹⁶ This is obviously problematic in cancer patients complaining of radicular pain in these distributions. Intrathecal administration at the appropriate level will result in a more predictable effect. For all the reasons enumerated above, we prefer intrathecal to epidural drug administration for the relief of cancer pain.

Mechanism of Intrathecal Drug Distribution

The importance of intrathecal volumes, versus that of the concentrations and doses, for the intensity and spread of spinal analgesia during chronic infusions of IT-morphine and bupivacaine is still unclear. Kroin et al.¹¹⁷ could not demonstrate any advantage in placing the catheter tip at more rostral locations, such as at the midthoracic or cervical cord when using hydrophilic drugs for pain relief. Van Zundert et al.¹¹⁸ demonstrated that a constant 70-mg dose of subarachnoid lidocaine delivered in a 0.5%. 1%, 2%, 5%, or 10% solution produced the same pinprick level of analgesia, degree of motor block, and duration of spinal anesthesia. Wagemans et al.⁶³ measured the pH of CSF in a patient receiving high doses of intrathecal morphine who failed to demonstrate either motor or sensory block after injection of 75 mg of lidocaine 5%. This patient’s CSF pH fell outside the physiologic range at 7.19 (normal range: 7.27 to 7.37), suggesting a possible explanation for the decreased activity of the local anesthetic. Kroin et al.¹¹⁹ noted an absence of ataxia at ITbupivacaine doses of 0.4 to 0.5 mg/hour in dogs. The human CSF volume is approximately 10 times that of a dog. Therefore, one might reasonably assume that doses of bupivacaine up to 3 mg/hour are safe in patients without evidence of neurologic deficit. When patients have neurologic deficits due to nerve damage or polyneuropathies, it is appropriate to reduce doses substantially (0.5 to 1.0 mg/hour) to avoid causing motor disturbances.

The location of the infusion catheter tip relative to the targeted spinal cord segments is critical because of the limited capacity for CSF to distribute drugs away from the
catheter tip. It is important to understand the mechanism of intrathecal drug delivery when administered as a slow intrathecal infusion. Human and animal studies of intrathecal drug distribution have been conducted as single bolus injections made at relatively rapid injection rates of tens of milliliters per minute.¹²⁰ However, infusion pumps used for the delivery of intrathecal morphine and bupivacaine deliver drug at much lower rates, sometimes as low as a few microliters per day. Bernards¹²⁰ characterized the distribution of baclofen and bupivacaine within the CSF and spinal cord at very low infusion rates in a pig model.¹²⁰ Bupivacaine concentration, when infused at 20 µL/hour, differed significantly between the anterior and posterior halves of the spinal cord and as a function of distance from the site of administration (Figs. 22.4 and 22.5). For both bupivacaine and baclofen, most of the drug recovered in the CSF and spinal cord was found within 1 cm of the site of administration. In the same experiment, a faster infusion rate of 1000 µL/hour imparted slight forward motion to the injectate with slightly better drug distribution. This suggests that there is poor mixing within the CSF and that drug distribution with slow continuous infusion is very limited. Of interest, baclofen, a more hydrophilic drug than bupivacaine, in bolus doses or infused at 1000 µL/hour distributed between the anterior and posterior spinal cord, but did not do so at infusion rates of 20 µL/hour. Hydrophilic drugs may have better distribution at bolus doses or higher infusion rates within the CSF.

The primary mechanisms responsible for drug movement within the CSF are Brownian motion (slight effect), kinetic energy imparted by the act of injection (slight effect when infused at slow rates), and CSF motion itself. Drug will move in the CSF as long as it is suspended in CSF, and its distribution will be limited only by the rate and extent of CSF motion. Motion of CSF is driven by cyclic expansion and contraction of the CSF vasculature during cardiac systole and diastole. The magnitude of motion is greatest in the upper cervical regions and decreases with distance from the cranial vault, becoming negligible at the level of the cauda equina.¹²⁰ Spinal CSF motion is also divided into three distinct channels, medioventral, dorsomedial, and lateral.¹²¹ These channels are distinct by the timing and magnitude of CSF oscillation. Within a given channel, the timing and magnitude of CSF motion is highly variable, with some areas along the cord devoid of either rostral or caudad motion. Also circumferential motion is not present between anterior and posterior sides of the spinal cord, presumably because of anatomical barriers (dentate ligament and spinal nerve roots) traversing
the lateral subarachnoid space. Factors limiting net CSF movement are summarized in Table 22.3. Placement of the infusion catheter tip at a precise level within the CSF is critical, especially if low infusion rates are utilized.

FIGURE 22.4 Average bupivacaine concentration in spinal cord specimens. Bupivacaine concentration differed significantly between the anterior and posterior halves of the spinal cord and as a function of distance from the site of administration. (Reused with permission from Bernards CM. Cerebrospinal fluid and spinal cord distribution of baclofen and bupivacaine during slow intrathecal infusion in pigs. Anesthesiology. 2006;105:169-178.)

FIGURE 22.5 Photographs of approximately 3-cm-long section of spinal cord in which methylene blue was added to drug infusion (20 µL/hour for 8 hours). There is heavy staining of the posterior segment of the spinal cord adjacent to the tip of the infusion catheter (A), minimal staining of the lateral surface of the spinal cord (B), and no visible staining of the anterior surface of the spinal cord (C). (Reused with permission from Bernards CM. Cerebrospinal fluid and spinal cord distribution of baclofen and bupivacaine during slow intrathecal infusion in pigs. Anesthesiology. 2006;105:169-178.)

TABLE 22.3 FACTORS LIMITING NET MOVEMENT WITHIN THE CEREBROSPINAL FLUID

Factor	Effect
Cardiac cycle	Propels CSF in opposite directions during each cycle
CSF pulse wave	The more caudad from foramen magnum, the smaller the pulse wave
Regional	Some areas are devoid of motion (at least sometimes)
Anatomical	Most CSF motion occurs in rostrocaudal axis, not circumferentially
CSF, Cerebrospinal fluid.

Clinical Efficacy of Intrathecal Local Anesthetics with Morphine

A number of clinical studies support the use of intrathecal bupivacaine and morphine for severe, intractable pain. Two reports, published in 1990 and 1991, first described the benefits of this approach in cases of refractory cancer pain.¹², ¹³ Nitescu et al.¹²

Only gold members can continue reading. Log In or Register to continue