Section 4: Subspecialty Management
Chapter 70: Chronic Pain

Nerve Blocks

Nerve blocks are often highly successful in controlling acute pain, such as that related to trauma and surgery. 3  Enthusiasm for the use of nerve block therapy in relation to chronic pain has waxed and waned as the role of interventional techniques has come under repeated scrutiny in the search for scientific documentation of undeniable benefits. Caution about the interpretation of nerve blocks traditionally viewed as clearly of value has been raised most recently by Hogan and Abram. 70  Their contention is that few studies exist to verify or to substantiate the diagnostic or prognostic value of most nerve blocks because of the vagaries of anatomy, the varying strengths of the solutions injected, and the bias the practitioner has toward achieving a positive benefit.

Chronic pain states have multiple causes, 1, 4, 5, 6, 7, 25, 71  and it is intuitive that nerve blocks have their greatest therapeutic value in pain problems in which nociceptive stimulation dominates. When chronic pain results from behavioral or psychiatric problems, a nerve block is unlikely to relieve symptoms because there will be multiple, more significant, nonsomatic contributions to the “pain,” and one should be concerned about the high possibility of inducing complications that are at least perceived to be related to nerve block therapy. Establishing the source of a pain generator is crucial to appropriate, effective pain treatment. Nerve blocks may be used diagnostically to determine whether an afferent nociceptive stimulus exists and perhaps can aid in establishing which neuropathways are involved. 72, 73  When such information is at hand, prognostic nerve blocks may predict whether surgical or neurolytic or thermal (radiofrequency [RF] or cryotherapy) interruption of the specific nerve path would be warranted or whether a series of therapeutic nerve blocks should be undertaken. These interventions are exceedingly common and run the gamut from simple infiltration of myofascial trigger points to injection of alcohol or phenol into a major nerve plexus for relief of pain, as in terminal cancer. 72, 73, 74  As the subspecialty of pain medicine grows, the proper place for nerve blocks in patient management will be elucidated, as will the risks of acquired infection for practitioners. 75  As more nonanesthesiologists enter the field, greater emphasis will be placed on the use of treatment modalities other than nerve blocks, a change that will enhance the likelihood that the necessary comparative studies will be done.

General Principles and Patient Assessment

A workup of the patient must precede the use of any nerve blocks, because it is necessary to determine that there are no major medical contraindications to the application of such therapies (including blood-borne infection, infection at the proposed needle insertion site, major anticoagulation, or the patient‘s or surrogate‘s refusal). 72  Even when these objections are not found, the crucial question to answer in preparation for nerve block therapy is whether a nerve block is appropriate for this patient at this time. Just because one can provide the therapy does not mean that it is indicated. The patient must provide written, informed consent following a concise presentation of the risks and benefits and the rationale for the planned procedure. 72  The ability to manage the side effects and complications of any blocks performed is requisite for the safe practice of regional analgesia in contemporary pain medicine.

Because it is of little use to perform diagnostic nerve blocks on patients who have no pain on the day of the procedure, the patient should affirm the intensity of the pain on a verbal 0 to 10 scale, in which 0 equals no pain and 10 is pain so severe that the patient is desperate. 7, 26  Alternately, the patient could make a mark on a VAS. The conventional medical examination preceding any nerve block must include a neurologic examination with particular attention to documenting the preexisting sensory and motor deficits. 26, 72  Often, very nonspecific information is obtained from patients undergoing diagnostic nerve blocks. 7  For example, the response to minor inconveniences (positioning, fitting of the blood pressure cuff, placing an intravenous catheter) gives an important perspective on the patient‘s demeanor. Some patients describe their pain in dramatic tones and as being of incredible severity, yet they manifest florid pain complaints on assertion of a 25-gauge needle. They may also claim that this pain is more severe than the chronic pain they hope to have relieved. Conversely, patients who tolerate quite uncomfortable needling procedures with great stoicism also provide insight into the reserved nature of their complaints. Relevant studies, such as radiographs, may need to be reviewed, as will the records of any preceding interventional therapy. Large doses of analgesic or sedative drugs may interfere with the patient‘s feedback about the immediate benefit of the nerve block or a painful complication, so careful consideration should be given to medicating such patients.

Assessment of the response to nerve blocks is both subjective and objective. The rating of the pain on a VAS should be repeated after the block is done. One must relate any reported pain relief to the onset of analgesia and recovery of function in the blocked nerves to the drugs used. The practitioner should check both the dermatomal and myotomal extent of the nerve block and should assess the impact on autonomic function. 72, 73  The extended response of the patient to the nerve block (in addition to his or her verbal report) should be further accentuated by appraisal of changes in the functional activity level, sleep patterns, gait, and medication requirements. The duration of side effects of the nerve blocks should be noted, and the positive effects should be correlated with the subjective and objective results. Repeated nerve blocks, using agents of different duration, will help to confirm the validity of initial impressions about their benefit and to dispel concerns regarding a placebo effect. 76  One must be careful in interpreting the duration of clinical improvement following even short-acting local anesthetics because decreasing the degree of CNS sensitization may result in pain relief that exceeds the expected duration of the drug used. 1, 2, 3, 8, 21  In addition, blocks performed with plain local anesthetics may have profound effects even though the actual period of pain impulse transmission blockade is relatively short because the patient sees, finally, that something really can be done about the pain. With the patient‘s attitude so affected, benefit from and compliance with other therapy increase. Nerve block therapy is rarely provided in isolation, so the patient‘s cooperation with other components of the treatment program must also be assessed.

Nerve blocks can confirm the presence or absence of nociceptive stimulation as the cause of pain and thus may give a clue as to whether pain originates from a peripheral site, the CNS, or psychologic/behavioral sources. 72, 73  Furthermore, nerve blocks can isolate the specific nociceptive pathway by showing that a specific nerve block, such as a median dorsal branch block for facet pain, 77  repeatedly abolishes pain. In this way, nerve blocks can be used in a prognostic fashion to suggest the effects of longer-term procedures such as RF denervation, cryotherapy, neurolytic blocks, or even surgery. Unfortunately, the satisfactory shorter-term effects of nerve blocks performed with local anesthetics are not always manifested when a more permanent nerve block procedure is done. Therefore, destructive nerve blocks are often reserved for carefully selected patients and those with terminal cancer. 74, 78 

The differential (or graduated) spinal anesthetic is of historical interest as a diagnostic tool in pain evaluations. This technique makes use of the fact that different concentrations of local anesthetics can almost selectively block sympathetic, somatic, and motor fibers. 7, 70, 73, 79  Investigators have used differential spinal techniques to diagnose common and obscure pain problems. On the other hand, differential spinal techniques are one of the best examples of the cautions raised by Hogan and Abram 70  concerning the absolute certainty of conclusions drawn from diagnostic and prognostic blocks.

Myofascial Pain and Fibromyalgia

Myofascial pain may be a primary cause of pain after injury to muscles, bones, or joints, or it may be a secondary consequence of pain from a distant site that causes subsequent postural alterations and stress/strain of muscles and supporting tissues over time. 80, 81, 82  Thus, myofascial pain may persist long after the original injury has healed and may be associated with referred pain. This condition must be distinguished from fibromyalgia; characteristic symptoms include complaints of widespread, chronic, nondermatomal musculoskeletal pain (described as being steady, deep, aching, and often associated with reflex muscle spasm), morning stiffness, decreased range of motion of joints, sleep disturbance, activity compromise, fatigue, and multiple areas that are tender. 80, 81, 82  This is a diagnosis made over time and generally without the aid of specific laboratory tests. The disease is characterized by exacerbations and remissions, and no single psychologic profile is associated with it. Criteria for diagnosis are published. 83 

Treatment is based on establishing a program of options and educating the patient about the mechanical factors that aggravate the symptoms, such as repetitive motions in unsteady positions. 81  Medications that decrease inflammation (NSAIDs), decrease muscle spasm (cyclobenzaprine, 10 mg tid; baclofen, 10–20 mg PO tid) and enhance stage IV, non–rapid eye movement (REM) sleep (amitriptyline, 25–100 mg PO qhs; sertraline 50–100 mg qd) are beneficial. Exercises that stretch muscles and then restore normal posture have the most lasting effect on the patient‘s life.

Myofascial Trigger Points

Trigger points are tender areas in the muscles or their supporting tissues that are believed to be caused by trauma from a specific accident or chronic occupational positioning, such as from typing or poor posture. 81  A trigger point can be found by applying pressure that reproduces the patient‘s pain, which may have a nondermatomal but consistent referral pattern. Several pain areas and trigger points may exist in the same patient, and there is great debate over how many trigger point injections should be provided at one visit and over time. On examination, 81  the painful areas have been described as feeling “rope-like.” A positive “jump sign” is said to be present when the trigger point is palpated and the patient “jumps” away from the pain.

Anesthesiologists are frequently asked to perform trigger point injections at the site of maximal tenderness. 7, 81  Dry needling techniques without injecting any drug are said to be successful in relieving pain, a finding suggesting that mechanical stimulation of the trigger point may be more important than the actual injectate. 84  As with most chronic pain states, the earlier such intervention takes place, the better the prognosis. Topical application of vapocoolant spray can precede the actual injection of local anesthetics such as 0.5 to 1 percent lidocaine or 0.125 to 0.25 percent bupivacaine, or saline, all with or without steroids. 81  Follow-up evaluation of the localized analgesic effect is necessary. Trigger point therapy is relatively benign and well tolerated. The ultimate goal is to assist the patient in achieving analgesia such that they can participate in active and passive physical therapy.

Sympathetic Nerve Blocks

Sympathetic nerve blocks have been a traditional modality for diagnosing and treating patients with RSD and causalgia, 70, 72, 73, 79  now referred to as CRPS type I and type II, respectively. 1  The global term sympathetically maintained pain (SMP) has been coined to classify these problems for which chronic pain appears to be associated with sympathetic nervous system dysfunction. 1  This title then refers to syndromes that characterize the effects on the body (usually on an extremity) that are thought to be mediated by sympathetic nervous system dysfunction after trauma such as fractures, lacerations, ligamentous strains/sprains, infections, and surgical incisions or resulting from visceral or CNS diseases. No single theory has exactly explained sympathetic nervous system dysfunction after such varied injuries, and the pathophysiology of SMP is poorly understood. Roberts 85  proposed that input from myelinated lowthreshold mechanoreceptors (LTM) sensitized WDR neurons in the spinal cord, and this triggered atypical CNS processing of noxious and non-noxious sensations (allodynia). Eventually, sympathetic nervous system efferent activity can stimulate the peripheral sensory receptors in the absence of cutaneous stimulation. Raja et al 86  postulated that norepinephrine at the site of injury activates a1 -adrenoreceptors on the nociceptive afferent fibers and/or causes the release of algesic amines (prostaglandin E, serotonin, and bradykinin) from reactive cells. There is evidence that nerve injury induces the acquisition of an abnormal excitatory response to the presence of norepinephrine in the periphery and at the level of the dorsal horn. This explains the lack of correlation between sympathetic tone and temperature, 87  why sympathetic blocks do not relieve all symptoms, and why sufficient blockade is so difficult to verify. 70  Harden et al 88  measured norepinephrine and epinephrine levels from the extremities of patients with SMP. Rather than reaffirming the traditional concept of sympathetic nervous system hyperactivity in the affected limb, they showed upregulation of peripheral nociceptors that manifested a pathologic response to circulating catecholamines. Other proposed mechanisms for SMP include activation of an inflammatory process in the sympathetic ganglion, the release of sympathotropic factors (such as nerve growth factor), increased vascular permeability, and a failure of the “usual” opioid modulation of regional sympathetic ganglia. 89, 90 

Clinically, the degree of inciting trauma bears no correlation with the severity of the SMP syndrome. RSD and causalgia are the two most common types of SMP. In RSD (CRPS-type 1), the history of a specific nerve injury is not always elicited (but tissue injury is common), and the original injury does not result in demonstrable neurologic deficits. The incidence of RSD following traumatic brain injury and stroke is in the 12 to 25 percent range, and the treating clinician should maintain a high index of suspicion when caring for such patients. 89  In causalgia (CRPS-type II), there is always at least partial nerve damage as the apparent inciting event. Common to the CRPS syndromes are burning pain, allodynia, and hyperesthesia/hyperpathia (which can be thought of as exquisite sensitivity to stimulation). Table 70–2 summarizes the pertinent historical findings, but much variation in patient description and physical presentation is common. The onset of pain after the injury is variable pain and can spread from an initial area of involvement in an extremity to the trunk and even to the contralateral limb. 7  Laboratory studies are frequently not necessary, and tests that detect alterations of blood flow may not be of essential use in the patient‘s workup or in the documentation of improvement with treatment. Sherman et al 87  showed that videothermography is not an appropriate tool to use alone for either singlesession diagnosis or multisession tracking of RSD. Because of the possible a-adrenergic chemosensitivity, 1–10 mg phentolamine given intravenously over 5 to 10 minutes may be useful in distinguishing SMP from neuropathic pain 86  and in predicting the usefulness of Bier block therapy. 91  A decrease in the VAS for both the ongoing rest pain and the evoked pain is necessary to view the test as predictive.

TABLE 70–2. History, Signs, and Symptoms of Complex Regional Pain Syndrome

The most common treatment for the sympathetically maintained component of the CRPS syndromes has been to provide interruption of the apparent pathologic somatic-sympathetic interaction by means of sympathetic blocks, although the absolute rationale for this approach is challenged at least by consideration of the contemporary mechanisms for SMP. 70, 72, 73, 74, 79  A positive response to a block may relieve only some of the pain and indicates that the patient responded at that time. Treating the cause of the CRPS does not constitute its management. The therapeutic or diagnostic effect of the specific techniques of sympathetic blockade should not be confounded by additional sensory or motor block.

Stellate Ganglion Blocks

Because humans do not possess a stellate ganglion per se, the more accurate anatomic term is “cervicothoracic sympathetic block.” 72  Up to 15 mL of local anesthetic is injected into the lower cervical sympathetic chain region at the C6 level. 7, 71, 72, 79  The caudad spread in the appropriate prevertebral fascial plane anesthetizes the lower cervical and upper thoracic sympathetic ganglia and effectively blocks transmission of impulses from the ganglia to the ipsilateral upper extremity. Ready et al 92  and Galindo 93  used a side-port needle for cervicothoracic block to ensure the stability of the needle and injection of drug into the proper tissue plane because in so doing one can keep the point of the needle in contact with the Chassaignac turbercle. Satisfactory sympathetic blockade results without the need to withdraw the needle from its bony end point.

Cervicothoracic blocks are usually performed using an anterior paratracheal approach with the patient in the supine position, but other techniques are described in Chapter 43 . The regional anatomy predicts the potential side effects and complications from both the needle and the drugs with these techniques. Spread of the solution into the groove between the esophagus and the trachea blocks the ipsilateral recurrent laryngeal nerve, leaving the patient with a hoarse voice for the duration of the local anesthetic effect. If the solution is administered deep to the prevertebral fascia, the local anesthetic will spread posteriorly and laterally and will involve the somatic components of the brachial plexus. Some or all of the roots of the brachial plexus may then be anesthetized. If local anesthetics are used, this is not a serious problem; however, if neurolytic agents are injected, this complication can be catastrophic.

A serious complication occurs when the local anesthetic solution is injected unexpectedly into the vertebral artery. The vertebral artery is posterior to the anterior tubercle of C6 and runs in the foramina transversaria in the transverse processes of the upper six cervical vertebrae. If the exploring needle passes between these processes and rests on the posterior rather than the anterior tubercle, withdrawal of the needle could leave the tip of the needle in the lumen of the vertebral artery. Small-volume injections (<1 mL) of local anesthetic solution into this vessel can produce convulsions. Therefore, careful aspiration is mandatory, and not more than 1 mL of local anesthetic solution should be administered as a test dose. 94  Treatment of a convulsion consists of oxygen by mask and/or positive-pressure ventilation plus the intravenous administration of a short-acting sedative-hypnotic, i.e., thiopental (Pentothal), midazolam, or propofol. Spread of the local anesthetic solution to the epidural and/or subarachnoid spaces producing profound anesthesia for a variable period of time is conceivable, although uncommon, as is motor blockade of the cervical plexus, leading to phrenic nerve paralysis.

Lumbar Sympathetic Blocks

As the sympathetic chain leaves the thoracic area, it lies alongside the lumbar vertebrae, anterior to the psoas major muscle and its fascia and posterior to the aorta on the left and the inferior vena cava on the right. Because of this positioning, the sympathetic chain can be successfully blocked from a posterior approach (Ch. 43).

Figure 70–1 shows the approach for a lumbar sympathetic block. 7  In the traditional method, three needles inserted at L2, L3, and L4 are advanced through the paravertebral space to the sympathetic chain. It is more common now to insert just one needle at the L2 or L3 level to produce an effective block. 7, 72, 73, 79, 95  This more lateral approach is often more comfortable for the patient (Fig. 70–2). In the lateral approach, the needle is inserted 8 to 10 cm lateral to the L2 spine. 95  Through a local anesthetic skin wheal, the 10-cm, 22-gauge needle is advanced at a 45-degree angle toward the vertebral body. Injection of a 1-mL bolus of local anesthetic relieves the discomfort felt by the patient as the needle passes through the lumbar fascia. As the needle advances through the psoas major muscle, paresthesia may occur in the distribution of the lumbar plexus and usually over the anterior aspect of the thigh (the genitofemoral nerve distribution). If this occurs, the needle should be redirected until only the vertebral body is contacted. The needle is then withdrawn slightly, and the angle progressively increased until the needle is advanced and slips just anterolateral to the vertebral body. At this point and after an aspiration test is negative, a test dose of local anesthetic with radiographic dye is administered. When accurate location of the needle is confirmed, a 10- to 15-mL bolus of local anesthetic (0.25% bupivacaine or 1% lidocaine) is injected. The patient maintains the prone position to prevent the local anesthetic from tracking back between the origins of the psoas muscle and into the paravertebral space where the lumbar somatic nerves could be anesthetized, thus confounding the effect of the trial block. Extensive infiltration of local anesthetic during needle insertion can anesthetize components of the lumbar plexus (usually the L2 and L3 dermatomes and/or myotomes in the anterior thigh). The use of higher concentrations of local anesthetic can produce motor block and impairment of the ability to walk, necessitating supervision of the patient until this dissipates.

Click thumbnail to see full size image
FIGURE 70–1 Lumbar sympathetic block. This figure shows the relation of the lumbar sympathetic chain on the anterolateral aspect of the lumbar vertebral bodies to the large vessels and somatic nerves. Inset diagram shows how the needle must traverse and, in its final position, be anterior to the psoas major muscle and its sheath to block the sympathetic chain satisfactorily. If injections are made within the psoas major muscle, its sheath will prevent diffusion of the drug to the sympathetic chain, and somatic nerve block of the lumbar plexus will result. The 10-cm needle is introduced at an angle of 45 degrees, approximately 8 cm lateral to the cephalad end of the L3 vertebral spine, as shown in Figure 70–4. The block can also be performed with a single needle at the L2 or L4 vertebral level.

Click thumbnail to see full size image
FIGURE 70–2 Needle positions for lumbar sympathetic block and celiac plexus block. Although the specific site of entry for the needle may vary, the intent is for the point of the needle to reach the target area of the anterolateral aspect of the L2, L3, or L4 vertebra for the lumbar sympathetic block. The celiac plexus block is usually a bilateral maneuver in which needles approach the celiac plexus, which lies anterior to the T12-L1 junction, from either side. The lateral approach to a celiac plexus block is shown on the left. The needle enters below the 12th rib. An approach angling cephalad ensures correct placement. For lumbar sympathetic block, the needle is inserted as shown on the right at a similar distance from the midline, that is, four finger-breadths. Here, however, the approach of the needle is more horizontal.

As in the cervicothoracic ganglion block, complications of the lumbar sympathetic block consist of spread of the solution to the neuraxis and somatic nerves (lumbar plexus) and unexpected injection of the solution into a blood vessel. 7, 72, 73, 79, 95  If the transverse process is mistakenly identified as the vertebral body, and a large volume of drug is injected then into the intervertebral foramen, an ipsilateral lumbar plexus block and/or epidural spread of the solution may occur. Therefore, concentrations of drugs that do not produce motor blockade should be used. If the needle is advanced so far that it is sitting in either the aorta or the inferior vena cava, injections of local anesthetic could easily produce systemic toxicity (Ch. 43).

Neurolytic lumbar sympathetic blockade 74, 78, 95  has been recommended for persistent ischemia of the lower extremity and some forms of persistent pain, because the afferent pathways involved in such problems are believed to pass through the lumbar sympathetic ganglia. For neurolysis with 6 to 10 percent aqueous phenol or 50 to 100 percent alcohol, the needle position must be confirmed radiographically prior to injection. Small quantities (2–4 mL) of drug should be injected through two needles placed at L3 and L4, as opposed to the single bolus injection technique used in diagnostic and other therapeutic blocks with local anesthetics, to decrease the dose of neurolytic drug placed at any one site. Minor groin anesthesia occurs in about 8 to 10 percent of these patients for up to 3 to 5 weeks (genitofemoral neuralgia).

With sympathetic block therapy, a staircase pattern of improvement is sought with serial blocks with local anesthetics, meaning that pain is reduced more and for a longer period of time for each block. Wang et al 96, 97  have shown that not all patients treated conservatively or with blocks improve. If benefits are not progressive but have definitive, albeit temporary effects, consideration may be given to RF ablation or thoracoscopic or surgical sympathectomy. 74  Brachial plexus blocks or continuous epidural blocks with solutions of local anesthetics with or without opioids may enhance pain relief and may allow for concurrent physical therapy. The therapeutic benefit of the blocks will be enhanced if physical therapy is provided to restore range of motion and to improve functional recovery of the affected extremity.

Intravenous Regional Sympathetic Blocks

In 1974, Hannington-Kiff 98  produced prolonged sympathetic blockade by the intravenous regional administration of guanethidine. He wanted to provide sympathetic blockade for weeks at a time and to avoid repeated cervicothoracic or lumbar sympathetic blocks using conventional needle techniques. Regional intravenous administration of guanethidine was found to produce a 3-day sympathetic block of dystrophic pain of the upper extremity, whereas a cervicothoracic block with bupivacaine lasted approximately only 10 hours. Bonelli et al 99  reported similar lengths of sympathetic blockade with guanethidine. In another study, 100  reported pain and blood flow were significantly lower, and skin temperature was significantly higher, when this block was compared with placebo injections of saline. Because injectable guanethidine is not available in North America, nor is its replacement, reserpine, other drugs have been sought. Ford et al 101  and Hord et al 102  reported on the use of bretylium in an intravenous regional (IVR) technique. Poplawski et al 103  reported using an IVR technique with local anesthetic and 80 mg methylprednisolone (Solu-Medrol), and Vanos et al 104  reported the successful addition of 60 mg ketorolac to the injectate.

Because anesthesiologists practice pain medicine while providing sympathetic blocks, they should be aware of additional modalities that will enhance the patient‘s response to treatment: ongoing physical therapy, transcutaneous electrical nerve stimulation (TENS) over vascular channels, NSAIDs, oral steroids, a- and b-adrenergic or calcium channel blocking medications, and self-regulation techniques. 105  Epidural clonidine infusion (10–50 mg/h) in patients who responded to a bolus injection was proposed by Rauck et al. 106  Neurostimulation techniques are showing success in some chronic cases, and implanted narcotic pumps are also finding utility in some selected cases. Topical treatment with Emla, clonidine, or capsaicin is being investigated further. The admixture between SMP syndromes and neuralgic pain suggests that some patients with chronic cases will need management as for neuralgic pain.

Celiac Plexus, Hypogastric Plexus, and Ganglion Impar Blocks

Celiac plexus block is another of the sympathetic block techniques that has been especially useful for patients with intractable pain due to cancer of the pancreas or other upper abdominal viscera. 72, 73, 74, 75, 78, 107, 108, 109  Successful block of the celiac plexus denervates the abdominal organs from the gastroesophageal junction to the splenic flexure of the large colon. Although permanent neurolytic blocks work well for cancer of the pancreas, they are much less successful for nonmalignant pain, such as that due to chronic pancreatitis, because the pain relief lasts only a few months. The technique, more fully described in Chapter 43, is strikingly similar to that for lumbar sympathetic block, except an additional 45-degree angle is used to place the needle at the level of L1 rather than at L2 or L3. Figure 70–2 shows the needle position for this block.

The most frequently encountered complication of celiac plexus block is postural hypotension. Because blockade of the vasoconstrictor fibers to the viscera hinders rapid compensation for changes in posture, blood pools in the viscera when the patient assumes the upright position. This problem is usually transient, lasting only a few days after neurolytic blockade. During this period, the patient must learn to assume the vertical position in a slow and deliberate fashion, because rapid movement can cause fainting. The spread of local anesthetic or neurolytic agent to somatic nerves is particularly pertinent because of the large volumes normally required for celiac plexus block. Spread of the agent to the upper lumbar nerves can impair motor power or hip flexion and the ability to walk. This complication is especially problematic if neurolytic agents are used with the technique in patients with nonmalignant disease. Therefore, neurolytic celiac plexus blocks require radiographic confirmation of correct needle placement before the injection of large quantities of neurolytic agents. Because of the proximity of the needle tip to the aorta and the inferior vena cava, aspiration tests and the administration of test doses are necessary to minimize the distinct risk of intravascular injection.

It is possible to perform celiac plexus blocks with alternative approaches to the classic prone position, such as with the patient in a decubitus position. 110  Montero-Matamala et al 111  claimed that the anterior approach is better tolerated by all patients, but in particular by terminally ill, heavily sedated patients and by all those who would have difficulty tolerating the lateral or prone positions. Ultrasound was used to locate the celiac arterial trunk. Claims are made that this technique is faster and more convenient than the lateral or posterior approaches and that a single-needle technique is sufficient. Jain and Ketchedjian 112  more recently wrote about the anterior approach. Boas 113  described a “retrocrural” technique for blocking the splanchnic nerves in an effort to improve the success with this block. Although these different approaches have been described, 114  there is no apparent difference with regard to pain relief. Most techniques appear to be safe when practiced appropriately, which usually means neurolytic procedures are done under radiologic control and with the use of a test dose with local anesthetic prior to the use of neurolytic agents through the same needle.

Sharfman and Walsh 115  challenged the efficacy of neurolytic blocks in patients with pancreatic cancer. The celiac plexus block is a percutaneous technique that interrupts afferent and efferent traffic in the celiac plexus. Brown et al 107  documented the definitive and sustained benefit of this technique when used in selected patients. They also showed that the incidence of side effects and complications was low, but not zero. Eisenberg et al 108  provided meta-analysis of the literature concerning neurolytic celiac plexus block. Eighty-nine percent of patients had good to excellent relief for the first 2 weeks after neurolytic celiac plexus block; partial to complete analgesia was maintained in about 90 percent of patients who were alive at 3 months, and 70 to 90 percent of patients surviving longer than 3 months continued to have partial or complete pain relief as well. The complication rate was 2 percent and included such incidents as intravascular injection, epidural or subarachnoid injection resulting in paraplegia, or needle trauma to the kidney, lung, or intestine. Side effects including localized pain at the injection site, diarrhea, and hypotension occurred as well.

There has been an awakening concerning the effectiveness of blocks of the sympathetic nervous system at many levels to provide cancer-related analgesia. An example is superior hypogastric plexus block. 116, 117, 118  This bilateral, retroperitoneal plexus is located at the L5-S1 area (the sacral promontory), close to the bifurcation of the common iliac vessels. It is the continuation of the celiac and lumbar sympathetic chains on each side of the vertebral column, and it innervates the pelvic viscera via the hypogastric nerves. The first description of blockade of this plexus was provided by Plancarte et al in 1990. 117  Twenty-eight patients with cancers of the cervix, prostate, or testicle who had had a positive response to local anesthetic blocks were given 10 percent aqueous phenol under radiographic guidance. A mean decrease in pain of 70 percent was achieved, and additional therapy increased the analgesia to 90 percent reduction of pain, with these benefits sustained until the patient‘s death over the ensuing 3 to 12 months in 26 of 28 patients. There were no complications reported, although the potential exists for injury to sacral nerves, bladder or bowel perforation, incontinence, and intravascular injection. Others who have reported benefits with this technique include deLeon-Casasola 116  and Waldman et al. 118 

The union of the bilateral sympathetic chains into a single, retroperitoneal plexus (the ganglion of Walther or the ganglion impar) is located anterior to the coccyx at the sacrococcygeal junction. This can be blocked with local anesthetic and then with neurolytic agents to denervate the lower pelvic structures and the perineum. Plancarte et al 119  presented the first use of this technique in 16 patients with advanced pelvic cancer. From 70 to 90 percent pain relief was achieved in all patients, with a remarkably low incidence of side effects and complications.

Back Pain and Epidural Injection of Steroids

Back pain is a most common and complex type of chronic pain. 14, 15, 16, 29, 77, 120, 121, 122, 123, 124  Many chronic back and extremity pain problems are thought to arise from musculoskeletal sources. A subset of patients with low back pain will have phospholipase A2 (PLA-2) leaked from the nucleus pulposus as a plausible explanation for their radicular pain. 125  Because it has been shown that corticosteroids can counteract the subsequent inflammatory reaction in the nerve roots and surrounding tissues, several investigators have injected steroids, with or without local anesthetic agents, into the epidural space. 126  The technique consists of injecting 40 to 80 mg methylprednisolone acetate (Depo-Medrol) or 25 to 50 mg triamcinolone diacetate (Aristocort) in 5 to 10 mL of local anesthetic or saline into the epidural space at the level of the suspected pathologic process. In the past, similar doses of steroids, but in smaller volumes of only 1 to 2 mL diluent, had been injected into the subarachnoid space when spinal pain was refractory to epidural administration. This procedure fell by the wayside years ago when concerns about the tissue toxicity of the corticosteroid preparations was paramount. 127, 128  Subsequent research did not identify that the commercial preparations of corticosteroids are indeed toxic when injected into the subarachnoid space. 129, 130  More to the point, when one understands that radicular pain is believed to originate from inflammatory reactions in the nerve roots in the epidural space, the logic of depositing drugs in the subarachnoid space to manage this pathologic process is lacking.

The significant analgesic effect of epidurally administered opiates in postsurgical and trauma-related pain resulted in their addition to corticosteroids given in the epidural space to treat chronic low back pain more thoroughly. Although initial reports were enthusiastic, subsequent studies were unable to duplicate those results. 131  Intraspinal opiates alone can provide significant pain relief for chronic low back pain caused by nociceptive stimulation from the low back area. 132  Therefore, some patients with chronic low back pain may benefit from ongoing therapy with continuousdelivery systems for administering long-term neuraxial narcotics. The most stable location for the long-term perispinal delivery of opiates is the subarachnoid space, and contemporary therapy includes an intrathecal catheter attached to an implanted pump (after a positive response to a trial of percutaneous intrathecal injections with opiates has been documented). 77 

The true place of steroids in subarachnoid and epidural analgesic blocks is still not completely clarified. 126  Controversy rages with regard to this widely used therapy. Specific guidelines and indications are awaited. Neither the optimal number nor the volume of injections is known, nor is it established whether corticosteroids should be injected with or without local anesthetics or other diluents. Benzon 133  reviewed this controversial topic more than 10 years ago, and many of the essential technique-based queries remain unanswered. Steroids are probably best reserved for instances in which conservative therapy for acute radicular back pain has been ineffective after 4 to 6 weeks or when the patient is suffering a flare-up of chronic back pain that has radicular features. 12, 29, 72, 73, 77, 79, 124, 126, 134  Steroids are used by some clinicians for intra-articular facet joint injections. 77 

Neurolytic Nerve Blocks

Intractable pain of malignant origin may warrant nerve destruction with neurolytic agents by procedures other than those mentioned earlier. 43, 47  This is most commonly considered when life expectancy is short and when chronic pain has a malignant source. It is fair to admit that clinicians still do not have all of the data relevant to deciding the exact indications for nerve blocks or which patients need them and when. Clinicians must carefully select patients for neurolytic procedures, by considering their coagulation and immune status, and must be technically meticulous so as not to obviate the advantages of nerve block therapy (outpatient procedures with less risk than surgery, repeatability, ready availability, and decrease in the need for other therapies 135, 136, 137, 138  ). Factors that should influence the decision to use neurolytic regional analgesic techniques include the patient‘s general medical condition, the location and rapidity of growth of the pain generator, the type of pain, the patient‘s life expectancy, the risks of the proposed procedure, the patient‘s tolerance for narcotics and previous conservative therapy measures, the response to a diagnostic block and the tolerance for any related side effects, and access to other specialists. The use of neurolytic blocks represents the ultimate in clinical judgment, rapport, and technical precision. Criteria that must be applied in consideration of such a procedure include a localized source of pain, pain that is diminished with a diagnostic block, pain that is poorly responsive to other more conservative therapy modalities, the absence of coagulopathy, the absence of localized infection or tumor at the proposed site of injection, the patient‘s tolerance for potential sensory/ motor/continence impairment, and the patient‘s understanding of the risks. 139 

Because peripheral nerve destruction with alcohol or phenol is frequently followed by denervation/dysesthetic (neuropathic) pain that may be as severe, if not worse, than the original pain, most anesthesiologists access the subarachnoid or epidural spaces for neurolytic injection. 140  The more extensive block provided by these routes of administration is often desirable because of possible growth of the tumor and a subsequent increase in nociceptive stimulation.

The relative merits of the different neurolytic agents are a matter of controversy. Alcohol and phenol are the most widely used agents. 7  The injection of alcohol is very painful, whereas that of phenol, which is usually mixed with saline or glycerin, is painless. The neurolytic effects of alcohol are more intense, and the effect with the block can be evaluated immediately. Phenol has a biphasic action, because it behaves both as a local anesthetic and as a neurolytic agent. Therefore, the extent of the block immediately after the procedure, which may reflect the local anesthetic action, decreases over the ensuing 24 hours to reveal a block of lesser extent. These blocks are not truly permanent; sensation and pain return within weeks or months. Therefore, these procedures are most often suggested to patients whose life expectancy is shorter than this time interval.

Somatic nerve blocks are performed with 50 to 100 percent alcohol, as are blocks of the sympathetic nerves. Alcohol is readily available and is easily stored. For peripheral nerves, 5 to 20 percent phenol is usually diluted with saline or water. For subarachnoid injection of phenol, it is mixed with glycerin (which makes its specific gravity greater than that of cerebrospinal fluid) for use in a hyperbaric technique.

Subarachnoid Neurolytic Nerve Blocks

Neurolytic subarachnoid blocks are best suited for patients with cancers involving the cranial nerves or for tumors involving somatic nerves lying between the limb plexus (i.e., tumors of the breast, chest, abdominal wall, or abdominal viscera). 140  For hypobaric subarachnoid neurolytic nerve blocks, the patient is positioned on the operating room table with the dermatomes to be blocked positioned uppermost. This configuration requires considerable finesse with positioning of the table and appropriate padding and other support to ensure patient comfort and stability during the block. If more than one spinal segment needs to be blocked, needles are inserted at the appropriate spinal levels. Small, discrete aliquots (0.10–0.25 mL) of absolute alcohol are injected at each level until the desired analgesic effect is obtained. It is usually safer not to try to spread the alcohol over more than a single spinal level. When the desired analgesic effect has been achieved and correlated with appropriate dermatomal analgesia, the patient is left in this position for at least 20 minutes for consolidation of the block.

Hyperbaric subarachnoid neurolytic block is performed with phenol mixed in 10 percent glycerin. This mixture is not commercially available and must be prepared immediately before the block by the hospital pharmacy department. 7  Because glycerin is very viscid, wide-bore needles (18–21 gauge) are used for injection. Positioning of the patient places the affected side down and the appropriate segmental levels in the most dependent position. The spinal puncture is performed at the appropriate level, and discrete 0.5 mL aliquots of the phenol-glycerin solution are injected until the desired pain relief and dermatomal analgesia have been obtained. As mentioned earlier, treatment of several segmental levels is accomplished more safely by using needles at each level rather than by injecting a larger bolus at a single level.

As an alternative to the traditional neurolytic agents alcohol and phenol, Korsten et al 141  demonstrated, in a small series of 12 terminally ill cancer patients with intractable pain, the effectiveness of a highly lipid-soluble congener of benzocaine, which appeared to afford satisfactory analgesia without creating compromise of motor or bladder/bowel function. Protocols for the clinical use of this agent are in development.

Epidural Neurolytic Nerve Blocks

Epidural nerve block is not widely performed using neurolytic agents. 72, 74, 140  However, epidural injection of 5 or 10 percent phenol in saline has been used successfully for some types of chronic cancer pain. This technique is particularly useful for bilateral pain and is not associated with major motor blockade; it therefore may be appropriate in the region of the limb plexus. It is difficult to detect any evidence of nerve deficits within a day or so of the block, although pain relief often persists considerably longer after this procedure. 7  This is a relatively safe neurolytic procedure, but it carries the theoretical risks of postinjection neuritis, excessive spread, and accidental subarachnoid placement with possible catastrophic paralysis.


The most frustrating problem with such procedures is usually failure to relieve pain after an apparent satisfactory block. 7  Involvement of nerve structures other than the intended neural elements can lead to unplanned sensory or motor deficits. 7, 74, 78, 107, 108, 109, 116, 140, 142  Thus, establishing the likely diagnosis for the cause of the pain and feeling confident that it can be approached with regional analgesic techniques will go a long way in avoiding these realities. In addition, many cancer pain patients use the intensity of their pain as an index of the activity of their disease. Thus, offers to eliminate the pain must be very clearly understood to guarantee satisfaction on both ends of the needle. Many cancer patients find comfort just in knowing what will be available when needed and will choose not to have all their pain relieved. One must be careful not to exchange tolerable pain for weakness, numbness, incontinence, and/or neuropathic pain. Sphincter disturbance is a very real concern with neurolytic blocks, especially those performed in the lumbar or caudal areas. 7, 140  Retention of urine with overflow incontinence is one of the most common complications. This condition is treated with an indwelling urinary catheter until such time as recovery occurs and the patient can be trained to urinate at regular intervals. These complications usually resolve with days to weeks. Obviously, patients must be thoroughly forewarned of the possibilities, informed consent must be obtained, and preexisting neural deficits must be documented. It is unlikely that a neurolytic block will eliminate all the patient‘s pain, even when the procedure is performed perfectly. Thus, a compassionate treatment program that blends medication and nonmedication modalities, as in hospice care, will still be needed 135, 136, 138, 143, 144 (Fig. 70–3).

Click thumbnail to see full size image
FIGURE 70–3 A daily diary kept by the patient can be very helpful in determining the nature and cause of chronic pain. The patient records all daily activity (sitting, walking/standing, and reclining). The right side of the diary documents the time and doses of medication; and the last column, the level of pain (on a scale of 0 to 10). If pain “switches off” at night, as in this example, pain may well be caused by environmental factors, rather than tissue damage. In this patient, the record of medication does not indicate drug dependence. Also significant is the inordinate amount of downtime. Specifically, the patient is active for only 2¼ hours a day, the rest of the time being spent reclining or sitting. This fact suggests that a behavior-modification program to improve the level of activity would be appropriate.

Neuraxial Administration of Narcotics and Other Drugs

The profound analgesia produced by administration of opioids at spinal cord receptors has led to tremendous interest in the clinical applications of this technique (Ch. 43). Wang et al 145  initially demonstrated therapeutic feasibility using subarachnoid injections of morphine in patients with terminal cancer. Since then, increasingly sophisticated delivery systems have been used for providing such therapy. 135, 136, 139, 144, 146, 147, 148  It is possible that the patient will have the postoperative epidural catheter left in place so that analgesia can be provided on a continuous basis into the extended in-hospital recovery period or even after discharge from the hospital. 135, 149, 150, 151  Narcotics delivered by indwelling catheters can provide analgesia for patients on an outpatient basis, 152  so patients with terminal cancer can now receive pain relief at home, where home health practitioners provide an increasing array of sophisticated care options. The ability to tunnel these catheters subcutaneously so that they emerge at an anterior abdominal wall site has helped to make this application possible. Automated external and implantable pumps are now also available. 146, 147, 148  Morphine has been the most widely used drug for epidural administration of narcotics for chronic pain. The side effects of its long-term administration (pruritus, urinary retention, mental status change, respiratory depression) occur only rarely and are manageable in patients with terminal cancer. 139, 143, 146, 147, 148 

Hogan et al 153  studied the use of epidural opioids and local anesthetics in referred patients with cancer pain. Of 1,205 patients, these investigators identified 16 who needed perispinal drugs. Six of the 16 patients obtained relief with just morphine, whereas the remaining 10 needed local anesthetic added. Common components of an epidural infusion could include opioids, local anesthetics, clonidine, and, occasionally, steroids. 146, 147, 148  Local anesthetics have been suggested as adjuncts in most infusions for chronic pain because one generally wants to avoid inducing weakness, numbness, and/or postural hypotension, which are the primary pharmacologic effects of the drugs. When patients show refractoriness to the opioid infusion, local anesthetics have been used to “rest” the opioid receptors and to restore their effectiveness during a brief hospitalization.

Detractions from epidural opioid infusions other than tolerance include the lack of expertise for initiation of such treatment and/or its maintenance in the patient‘s home area, the labor intensiveness of the follow-up care, and the cost. Sjoberg et al 156, 157  provided an important study that addressed the safety of long-term intrathecal injections of drugs for the treatment of refractory cancer pain. They detailed the neuropathic changes in 15 patients with infusions of morphine and bupivacaine for a median of 81 days (range, 4–274 d). 156  No patient had neuropathic changes correlated with the duration or cumulative doses of the intrathecal therapy. This group also refined the recommendations about specific drugs one could infuse in patients with intractable cancer pain. 157  Fifty-three patients had VAS scores decreased from 6 to 7 of 10 to 0 to 2 of 10 during an infusion duration of 7 to 334 days while receiving a continuous infusion of morphine and bupivacaine in roughly a 1:10 ratio. Reported side effects were associated with bupivacaine and included urinary retention, paresthesias, gait impairment, and, occasionally, orthostatic hypotension.

The search for “new” analgesics will continue. DuPen et al 150  demonstrated the feasibility of combining morphine with low-dose bupivacaine to maximize analgesia even in a home-based setting. Steroids may be beneficial in decreasing the consequences of inflammatory neuropathy, as when cancer invades neural structures in the epidural space. Clonidine, an a2 -adrenergic agonist, has become available for general use as a modality in neuropathic pain control. 154, 155  Through actions at presynaptic and postsynaptic a2 -receptors of the spinal cord, 30 mg hourly can augment opioid-based analgesia with a low profile of side effects. 146, 147, 148, 154, 155  A central effect of benzodiazepines that is less dramatic for C fiber stimulation than it is for A-delta stimulation has been demonstrated, so these drugs may find utility in the future, as may NMDA antagonists, ion channel blockers, NSAIDs, and cholinesterase inhibitors. 147  Even with these valuable additions to the pharmacologic armamentarium, more modalities must be at hand. Unique techniques for regional analgesia other than neuraxial blockers, 158, 159  neuroablative procedures, 160  and neurosurgical treatments 161, 162, 163  represent such diversity.

Anesthesiologists have guided the subspecialty of pain management for years and have advocated the rational administration of drugs and the performance of nerve blocks to treat chronic pain. They have recognized, too, that other therapies are necessary. Because the treatment of patients with chronic pain has now become the practice of pain medicine, all health care workers must acknowledge the need for a treatment program, one that incorporates a number of therapeutic modalities (in addition to medications) used concurrently, regards all the discovered contributors to the “pain,” and fosters routine follow-up so that the therapeutic plan can be modified to include only those treatments that are contributing positively to the patient‘s quality of life.