Recent advances in the management of oncologic pain

RECENT

ADVANCES IN THE MANAGEMENT OF ONCOLOGIC PAIN

Over the past decade, extensive advances have been made in pain management and related fields. Strides have been greatest in the related areas of conceptual and practical knowledge, whereas widespread improvement in actual delivery of care is only now forthcoming. The discovery of endogenous opioid receptors has laid a sound theoretical basis for the neuroaxial administration of exogenous opioids. Observations on the role of nonopioid substances in pain transmission have stimulated investigators to initiate trials of spinal cord analgesia with substances as diverse as clonidine,’ calcitonin,’ and somatostatin.3 The principles of the gate control theory have been utilized to provide pain relief to hundreds of thousands of patients in the form of portable transcutaneous electrical nerve stimulation units. Improvements in the technology of miniaturization have increased the applications for spinal cord stimulation by permitting lead implantation through an epidural needle, in most cases eliminating the need for laminectomy. Deep brain stimulation, still in the early stages of development, holds great promise for the provision of widespread pain relief with minimal potential for physiologic and psychological dysfunction. Palliative radiotherapy and chemotherapy are being used earlier, with increased frequency. New formulations of analgesic drugs have been introduced, and rational pharmacologic principles have been proposed and accepted.4 Pharmacologic intervention, with reliance on the oral route, continues to be the mainstay of care for most patients with cancer pain. Effective drug therapy is facilitated by the use of new controlled release preparations of oral opioids and increased utilization of an around-the-clock dosing schedule supplemented by “rescue” doses of short-acting drugs. Nerve blocks continue to play an important role in the treatment of intractable cancer pain and remain a primary focus of the anesthesiologist. Techniques have been refined, and as with other modalities that have been utilized with partial success, the role of neural blockade is now complemented by the support offered by the multidisciplinary pain management team. The promulgation of a multispecialty framework for comprehensive assessment and delivery of care is one of the most important developCut-r

Prabl

Cancer,

May/June

1989

135

ments in contemporary pain management. The visionary multidisciplinary pain center, run by any of a number of specialists, involves the input of psychiatry or psychology, anesthesiology, neurology, neurosurgery, dentistry, oncology, rehabilitation medicine, physical therapy, and nursing. The mechanics of the delivery of care have not kept pace with technological and conceptual developments. The availability of narcotic drugs, skilled physicians, and technological support is inadequate in many parts of the world, including areas in the United States. Fear of drug addiction by patients, their families, physicians, and policymakers has long been an impediment to optimal usage of potent narcotics for pain control. Comprehensive pain control programs, particularly for preterminal patients, are costly in terms of dollars, time, and effort. Philosophical and economic commitment at institutional, departmental, and individual levels are essential for the establishment of successful programs. The horizon is bright for the next decade. Assuming the continued intensity of current research, further technical and conceptual advances can be expected. More important from practical and humanitarian aspects, corresponding increases in both the quantity and quality of care delivered can be predicted based on a number of factors. Recognition of the problem of inadequately treated cancer pain is increasing owing to initiatives by the medical-scientific community and public interest groups, notably the World Health Organization WHO). Assistant U.S. Surgeon General John C. Duffy sums up the dilemma facing our health care system in his statement that “We know how to relieve cancer pain. This is not something to be looked into. It is, for us, the moral imperative to act.“5 Also, Director of the National Institute on Drug Abuse, Charles Schuster, comments that “drug addiction has nothing whatsoever to do with cancer pain.‘15 These statements serve as an official confirmation of the results of numerous scientific studies.6’ 7 The future of our ability to deliver care to patients with pain problems rests, in large part, on the adequate dissemination of knowledge. Once woefully inadequate, the quality and quantity of communication is ever increasing. Practitioners’ access to basic and clinical scientific advances is receiving new support through the recent development of specialty journals, the publication of new textbooks, and the convocation of regional, national, and international symposiums. The activities of the International Association for the Study of Pain (IASP), the American Pain Society (APS), and WHO have been instrumental in monitoring health care needs, setting standards of care, and supporting educational programs. The reader is referred to the IASP’s landmark publication, “Classification of Chronic Pain,“’ which was devised to promote the acceptance of a common taxonomy (definition of terms, classification and coding of 136

Curr

Probl

Cancer,

May/June

1989

syndromes). A further tion of state-of-the-art

example of organized, intentional disseminaknowledge is the publication of Principles of

Analgesic Use in the Treatment of Acute Pain and Chronic Cancer Pain,4 a reference suitable for use by algologists, other specialists, general practitioners, and nurses, developed by the APS with the support of the International Pain Foundation (IPF). The IPF is a nonprofit organization founded in 1986 to promote education and research. A committee of the IASP has recently completed a survey of postgraduate and undergraduate teaching related to pain and is formulating recommendations for future curricu1um.g The American Board of Anesthesiology now recommends exposure to formal teaching in pain management for all residents in anesthesia. DEFINITION

AND

PREVALENCE

OF CANCER

PAIN

The definition of pain proposed by the IASP, “an unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage”’ reflects current thinking, which recognizes pain as a complex, multidimensional experience. The concept of “total pain,” developed by Saunders,l’ goes further to recognize disturbances of mood, emotion, sexuality, appetite, sleep, posture, and activity as elements of the experience of pain. Prevalence of pain in patients with cancer has been demonstrated repeatedly.‘1’12 A recent analysis of 47 published reports revealed that 71% of patients with advanced cancer had pain as a major symptom and noted a 50% incidence of pain in patients with intermediate stage disease.13 According to the conservative estimates of WHO experts, over 3.5 million people suffer from cancer pain daily.14 Cancer refers generically to a multitude of different disorders that share certain common histopathologic features. Likewise, the term cancer pain describes a heterogeneous group of syndromes of diverse, often multifactorial etiology. The additional factor of subjective response to pain guarantees that the clinician will be exposed to symptoms that vary widely in severity and ease of treatment. Health care providers and the public have a tendency to view cancer pain as being uniformly severe,‘5”6 but this is not always borne out in experience. Daut and Cleeland” compared reports of pain severity in patients with metastatic cancer, arthritis, and chronic stable benign pain and found that cancer patients rated their worst pain at about the same level as did patients with rheumatoid arthritis. Both groups rated their worst pain at a significantly lower level than did patients in the chronic nonmalignant pain gro~p.~’ Conversely, even pain of mild or moderate intensity may be associated with significant distress when symptoms are related to cancer. These findings emphasize the importance of individual patient assessment. Cum Probl

Cancer,

May/June

1989

137

PATIENT

ASSESSMENT

The evaluation of the patient presenting with cancer-related pain serves multiple purposes. The initial encounter should be broadly based: rather than limiting inquiry to the pain syndrome per se, the process should encompass evaluation of the person, their feelings and attitudes about pain and disease, and family concerns. A compassionate but objective approach to assessment serves to instill confidence in the patient and family that will be of value throughout treatment. Thorough review of the patient’s records and a detailed pain history serve both to delineate the source of pain and to distinguish the degree to which the patient’s complaints are related to physical pain vs. psychic suffering. The referring physician who has known the patient over time is a source of valuable information and should be consulted personally. A standard text can be consulted for techniques of obtaining a comprehensive pain history.‘7’18 Psychological testing is of value, although in selected cases it must be abbreviated in consideration of poor physical and emotional condition.” Since clinical pain is regarded as a subjective experience, quantitative measurement necessarily depends on subjective tools. The tool used most frequently is the unidimensional visual analogue scale, which commonly consists of a horizontal line evenly “anchored” or divided into ten segments numbered consecutively between 0 and 10. Patients are told that 0 represents total absence of pain and 10 denotes the most severe pain imaginable, and are instructed simply to mark the number that best describes the level of pain they are experiencing. Modified visual analogue scales for use with children may substitute a continuum of smiling to crying faces for numbers or may utilize colors or a pain “thermometer.“20 The McGill Pain Questionnaire provides additional information. It employs a five-point scale labeled “no pain,” “mild pain,” “moderate pain,” “ severe pain,” and “intolerable pain,” as well as instructions to select from a list of 78 sensory and emotional adjectives commonly used to describe pain, and graphic representations of the body to be shaded to indicate where pain is located. A discussion of the validity, reliability, and reproducibility of these and other devices is beyond the scope of this chapter. Diagnostic maneuvers are pursued to determine the underlying cause of pain so that, when possible, therapeutic strategies can be directed at alleviating the source of pain in preference to relying on a purely symptomatic approach to treatment. Direction of treatment at the source of pain permits more focused care, reducing the likelihood of complications from overtreatment of symptoms. Identification of psychological causes of pain may permit the substitution of behavioral intervention or counseling for drug therapy. Escala13s

Curr

Probl

Cancer,

May/June

1989

tions in pain usually signal progressive injury to tissue, and anatomic localization of a painful lesion through radiologic studies or diagnostic nerve blocks may suggest additional treatment options. For example, the identification of spinal cord impingement or impending structural collapse may warrant the need for mechanical or surgical stabilization to forestall paraplegia. Pain of unknown etiology may be an early symptom of recurrent disease or undiagnosed malignant neoplasm. This is particularly true for herpes zoster. The pain consultant should not be dissuaded from ordering additional tests because the patient has already been “worked up” or because antineoplastic treatment has been abandoned. Nevertheless, often only symptomatic treatment is warranted, particularly when expectancy of life is limited. A number of schemata for classification of cancer pain have been suggested, the application of which can prove useful when considering diagnosis and treatment. One such classification is based on chronic@. Acute pain most often occurs in association with the onset of disease and is often associated with signs of sympathetic hyperactivity and heightened distress. Analgesics may be required on a transient basis, but symptoms often resolve as antitumor therapy progresses. The patient with chronic pain shows signs of biologic and psychological adjustment, and as a result, treatment is more complex. Classification based on intensity of pain determines where the patient is likely to fall along the “analgesic stepladder.” Patients can be broadly classified according to the extent to which their suffering is related to psychic vs. organic processes. A general classification by pathophysiology distinguishes somatic, visceral, and deafferentation pain, each of which has different characteristics and requires different approaches to treatment (see “Types of Pain” section). Foley has devised a classification of pain that takes into account patient characteristics and stage of disease, which is summarized in Table 1. Finally, the patient’s history, physical findings, and the results of radiologic studies aid the practitioner in determining the specific pathologic process that is present. Numerous distinct cancer pain syndromes have been recognized and described.‘g’22 Mechanisms include obstruction of lymphatic and vascular channels, distention of a hollow viscus, edema, and tissue inflammation or necrosis. Severe symptoms are most often related to direct invasion of pain-sensitive structures by tumor mass. Infiltration of bone is the most common cause of pain. Since up to 50% decalcification must be present before osseous lesions are visible on plain roentgenograms?3 scintigraphy (isotope scannin 1 is preferable to plain films for detecting most bone metastases. zf In contrast, primary bone tumors are well visualized with conventional radiography. Since osseous involvement in thyroid cancer and multiple myeloma rarely produces inCurr

Probl

Cancer,

May/June

1989

139

TABLE 1. Classification

of Cancer

Pain by Patient

Pain Syndrome Acute cancer-related Related to diagnosis Related Chronic

pain

to treatment cancer-related

Associated

with

pain treatment

Associated with progression Patients with preexisting chronic pain Patients abuse Dying

*Modified

with

history

of drug

Foley

KM: Treatment

patients

fmm

Type*

Patient

Characteristics

Patients tend to be hopeful Endure pain readily, often without seeking treatment Recurrence of pain can be devastating (identified with recrudescence of disease) Psychological adaptation/maladaptation established Disease quiescent; overriding concern is reestablishment of functional life-style Hopelessness, helplessness often predominate Require intensive intervention and support Pain behavior established Accurate diagnosis essential Difficult to evaluate and treat Risk of inadequate treatment Coordinate rehabilitation, social work Adequacy of treatment has impact on patient and family Assure comfort at all costs of cancer

pain. N Engl J Med

1985; 313:8&95.

creased uptake of isotope, conventional radiographs, computed tomography (CT), or magnetic resonance imaging are preferred.25 Because abnormal findings on scintigrams are not specific for malignant disease, they must be interpreted together with other radiologic studies and in the context of clinical findings. Neoplastic involvement must be differentiated from changes related to infection, trauma, or degeneration because treatment differs. Although the majority of skeletal metastases do not produce painF4 pain from bony metastases can present with a wide variety of symptoms. Pain is usually constant but may be greatest at night and is often worse with movement or weight bearing. Patients may report a dull ache or deep, intense pain, and there may be referred pain, muscle spasm, or paroxysms of stabbing pain, particularly when bony lesions are accompanied by nerve compression. Knee pain associated with metastatic involvement of the hip is an important example of referred pain. Specific syndromes associated with bony spread of tumor to the base of the skull have been well described by Foley.” Jugular-foramen syndrome is usually associated with occipital pain, often radiating to the vertex or ipsilateral shoulder, and may be accompanied by local tenderness and exacerbation with movement of the head. Neurologic signs consistent with dysfunction of cranial nerves IX through XII and Horner’s syndrome may be present. Clivus me140

Curr

Probl

Cancer,

May/June

1989

tastases are associated with vertex headache exacerbated by neck flexion and may be accompanied by either unilateral or bilateral lower cranial nerve dysfunction WI through XII). Sphenoid sinus metastases are suggested by severe bifrontal headache radiating to both temples with intermittent retro-orbital pain. There may be complaints of nasal stuffiness and diplopia, and a unilateral or bilateral sixth-nerve palsy may be present. Radiologic documentation of base of the skull syndromes is difficult, and workup usually includes lateral roentgenogram of the neck, tomograms, and CT scan of the head. Meningeal carcinomatosis is a syndrome of tumorous infiltration of the meninges, usually with parenchymal involvement. Typical meningeal signs may be present, as well as deficits in cranial or upper cervical nerve function and alterations in mentation. Pain occurs in 40% of patients, either as constant headache and neck stiffness or diffuse lower backache.” Lumbar puncture is usually diagnostic. Malignant involvement of the vertebral column is frequently associated with localized spinal an&or radicular pain and often weakness and tingling in a dermatomal pattern. Rapid progression of neurologic deficit, particularly of motor weakness, or the appearance of urinary or fecal incontinence indicates progressive epiduralspinal cord compression and warrants urgent intervention. In a review of 130 cases of spinal cord compression, 96% of patients were found to have increased pain as their presenting symptom, and in 10% of patients, pain was the only abnormal clinical finding.z6 Generally, myelography or CT scanning should be carried out, and treatment is either with immobilization and high doses of corticosteroids and/or radiation, or decompressive laminectomy. Surgery is not advisable when overall clinical condition is poor or spinal metastases are present at multiple levels. Though invasion of the upper lumbar vertebrae is usually heralded by dull backache and radicular signs, it should be noted that pain may be localized to the sacroiliac joints or iliac crests, and that radiologic investigations of the lumbar spine should not be overlooked when these symptoms are present.” Invasion or compression of somatic nerves by tumor is generally associated with constant, burning dysesthetic pain, often with an intermittent laminating component. Diffuse hyperesthesia and localized paresthesias are not uncommon, and muscle weakness and atrophy may be present if the affected structure is a mixed or motor nerve bundle. Horner’s syndrome may be present with brachial plexus involvement (Pancoast’s or superior sulcus syndrome). Suspected brachioplexopathy requires a complete radiologic evaluation, which may include myelography if epidural extension seems likely. Invasion of the lumbar plexus is associated with radicular pain or referred lower-extremity pain. Radiologic studies may not corroborate clinical findings,” and diagnostic nerve blocks can be quite helpCur-r Pmbl

Cancer,

May/June

1989

141

ful. Invasion of the sacral plexus is most often associated with severe constant lower backache, often with progression to include perineal sensory loss and bowel and bladder dysfunction. Plain films, tomography, and scintigrams frequently demonstrate bony invasion of the sacral plates. Visceral pain is usually diffuse and vague and may be referred to distant sites. Painful stimuli include ischemia, chemical damage, spasm, overdistention of a hollow viscus, and capsular stretch. Referred pain is felt in the cutaneous area corresponding to the dorsal horn neurons on which visceral afferents converge and may be accompanied by localized allodynia and hyperalgesia, nausea, or fatigue.27 Pain may be constant if it is due to chemical irritation or invasion, or colicky, as in obstruction of the gastrointestinal, urinary, or biliary tracts. Local anesthetic blocks will often differentiate visceral from somatic pain. The syndromes mentioned above relate primarily to cancer progression, but as the literature points out,” pain associated with cancer therapy is also of great significance. These syndromes include postoperative pain, such as that following mastectomy, radical neck dissection, thoracotomy, and limb amputation. Sequelae to radiotherapy include spinal cord myelopathy, fibrosis of major nerve plexuses, osteonecrosis, mucositis, and radiation-induced peripheral nerve tumors. Vincristine and vinblastine chemotherapy may induce painful peripheral polyneuropathy, and aseptic femoral or humeral head necrosis may follow high-dose steroid treatment. Postherpetic neuralgia is not uncommon. Finally, when formulating a differential diagnosis, the clinician must be alert to nonmalignant causes of pain. PHARMACOLOGIC

MANAGEMENT

An estimated 85% to 90% of patients with cancer-related pain experience effective pain relief with appropriate pharmacologic management.” If this estimate seems high from the anesthesiologist’s perspective, it is in part because referral is usually based on failure of conventional drug management. Nevertheless, current utilization of pharmacologic therapy is still suboptimal; the results of an analysis of 12 published reports describing the management of over 2,500 cancer patients in developed countries revealed an absence of satisfactoiy pain relief in 50% to 80% of caseszg

NONNARCOTICANALGESICS These drugs are effective as initial agents for the treatment of various acute and chronic pain conditions, including pain associated 142

Curr

mob1

Cancer,

May/June

1989

with arthritis, trauma, and cancer. Less potent analgesics are often ignored in favor of narcotics for the treatment of cancer pain because of the mistaken assumption that all cancer pain is equally severe?’ Likewise, when cancer pain is refractory to nonnarcotic analgesics, there is a tendency to replace these drugs with narcotics, ignoring the possibility of additive or synergistic analgesia.3o Though neither class of agents is free of side effects, the concept of reducing narcotic requirements by the continued administration of nonopiate analgesics is effective in limiting the adverse effects associated with higher doses of narcotics, as well as delaying the development of tolerance. It is of note that, at usual doses, the “mild narcotics” kodeine, dihydrocodeine, propoxyphene, pentazocine) have not been shown to be more analgesic than aspirin or acetaminophen, and that the mild narcotics are most effective when administered concomitantly with aspirin, acetaminophen, or nonsteroidal anti-inflammatory agents.31 The nonnarcotic analgesics are a heterogeneous group of drugs often considered together because of common indications for their use rather than similarity of structure or uniformity of pharmacologic action.31 Of the nonnarcotic analgesics, the nonsteroidal antiinflammatory drugs (NSAIDs) are particularly useful because, in addition to providing pain relief, per se, these agents reduce stiffness, swelling, and tenderness. Simple pain relief accompanies intermittent use. A few days of regular use are required before their antiinflammatory actions exert themselves. A ceiling effect has been demonstrated, above which no further analgesia can be expected, but the dosage at which this is observed clinically varies among patients. Tolerance and physical dependence do not occur. A variety of NSAIDs are available. Since most clinicians believe that these drugs are interchangeable, the most useful classification may be on the basis of side effects and toxicity (Table 21. The common problems associated with the use of NSAIDs in cancer patients are related to gastric irritation, prolonged bleeding times, and masking of fever. Phenylbutazone should be avoided because of the uncommon but grave occurrence of agranulocytosis and aplastic anemia. The NSAIDs, as well as steroids, are particularly effective for the treatment of metastatic bone pain. A biochemical mechanism is attractive to explain how even very small bony metastases can be associated with severe pain. The NSAIDs act peripherally, probably reducing pain and inflammation through inhibition of the cyclooxygenase pathway of arachidonic acid breakdown3’ Inhibition of the enzyme cycle-oxygenase decreases the formation of prostaglandin 2 (PGEJ. Osseous metastases elaborate PGE2F3 which probably contributes to pain by sensitization of peripheral nociceptors to various substances, including bradykinin, histamine, and 5-hydroxytryptamine.34 As deposits enlarge, stretching of periosteum, pathoCurr

Probl

Cancer,

May/June

1989

143

TABLE Classification

2. of NSAIDs

by Toxicity*

Associated with few GI side effects Fenbufen Ibuprofen Naproxen Anything but suppository Not aspirin Not indomethacin Not flurbiprofen Preferable in patients receiving anticoagulants Indomethacin Naproxen Ibuprofen Piroxicam Not aspirin Not phenylbutazone Monitor carefully More likely to cause rashes Fenbufen Fenclofenac Feprazone Suitable for children Aspirin Ibuprofen Naproxen Mefenamic acid ‘Modified fmm Huskisson EC: Non-narcotic analgesics, in Wall PD, Melzack R teds): Textbook of Pain, New York, Churchill Livingstone, Inc. 1984, p 510.

logic fracture, and perineural invasion contribute requirements for more potent analgesics increase. NARCOTIC

to pain,

and

ANALGESICS

Narcotic analgesics are the mainstay of therapy for the control of moderate to severe pain in patients with cancer. The medications in current use do not differ significantly from those previously available, although new guidelines for administration have been developed and prescribing patterns have changed. The reader is referred to other sources4’14’21 for descriptions of the pharmacologic profiles of the useful narcotic analgesics. The following guidelines summarize recent suggestions for the rational pharmacologic management of cancer pain. 1. Comprehensive assessment ideally precedes initiation of therapy. Treatment strategy should be directed toward relief of “total 144

Cur-r

Pmbl

Cancer,

May/June

1989

Concomitant functional disability (disturbance of sleep, appetite, mood, activity, and sexual function) should be determined and addressed by the treatment plan. Concrete goals, acceptable to the patient, family, and treating physicians, should be established. 2. As with less potent analgesics, narcotic therapy should be individualized or “tailored” to suit the patient’s needsT5 Dose response and side effects vary widely based on a number of physiologic and behavioral factors.36S37 3. Once an acceptable drug regimen has been established, adequacy should periodically be reassessed. Patients are often reluctant to request more potent analgesia. Increased drug requirements related to progression of disease and development of physical tolerance should be anticipated. Tolerance is most frequently manifested by decreased duration of analgesic effect. 4. When NSAIDs provide insufficient relief of pain, or are poorly tolerated, the addition or substitution of a codeine preparation is usually recommended as an analgesic of intermediate potency. 5. When combinations of codeine and NSAIDs or codeine alone are insufficient, therapy should progress to include more potent narcotic analgesics, in a “stepladder” fashion3’ (Fig 1). Less potent analgesics should not be summarily excluded, since NSAIDs may

p*.“‘O

Pain Pain persisting or increasing

persisting

of

inc-r

IIFrwdom ’ llwin

FIG 1. World (From Curr

Health Organization-advocated World Health Organization:

Probl

Cancer,

May/June

X389

Cancer

stepladder approach Pain Relief. Geneva,

toward cancer pain relief. 1986. Used by permission.) 145

provide additive or synergistic analgesia, and codeine preparations may be useful for breakthrough or incident pain. When a narcotic is initially introduced, the starting dose should be low since early appearance of side effects will influence compliance. 6. Patient and family education is an essential element of a successful pain relief program and is ideally accomplished through the combined efforts of physicians and nurses. Patients frequently maintain deeply rooted cultural fears of addiction. The distinction among psychological addiction, physical dependence, and tolerance should be explained. Concepts and details of side effects, tolerance, overdosage, and withdrawal should be explored. Treatment strategy should anticipate and avoid these problems. 7. A time-contingent analgesic schedule is more effective than symptom-contingent administration; if analgesics are witheld until pain becomes severe, sympathetic arousal occurs and even potent analgesics may be ineffective. Established patterns of anticipation and memory of pain contribute to suffering, even during periods of adequate analgesia. Around-the-clock administration of appropriate analgesics maintains therapeutic blood levels and decreases the likelihood of intolerable pain.3s Patient compliance is enhanced by the prescription of long-acting agents (levorphanol, methadone) or slow-release formulations of morphine (MS Contin, Roxanol). 8. Selected patients benefit from the addition of adjunctive drugs, including tricyclic antidepressants, antihistamines, amphetamines, phenothiazines, anticonvulsants, and steroids. Almost all patients receiving opioid therapy will require a laxative, and about two thirds will benefit from the regular administration of an antiemetic.40 9. When possible, analgesics should be administered orally to promote independence and mobility. 10. When pain control is inadequate with analgesics taken by the oral route, or the oral route is contraindicated, consideration should be given to alternative means of drug delivery. Oxymorphone hydrochloride (Numorphan) rectal suppositories provide 4 to 6 hours of potent analgesia. Other options include continuous subcutaneous infusion by means of a portable drug pump, patientcontrolled analgesia (intravenous or subcutaneous), and intrathecal or epidural opiates administered via an externalized catheter or internalized pump. New means of administering narcotics are being explored and include mucous membrane (sublingual and intranasal) and transdermal absorption. A report appearing in JAMA describing the successful bedside administration of nitrous oxide to a series of dying adolescents is a further example of innovative care4’ 11. Maintain familiarity with the pharmacologic profiles of a variety of analgesics. Consider drug substitution when a patient exhibits tolerance to the analgesic effects or intolerance to the 140

Curr

Probl

Cancer,

May/June

1989

side effects of medications. Half the calculated equianalgesic dose of the new drug is recommended as a starting dose, which is then titrated upward as needed’l (Table 3). 12. Avoid chronic administration of meperidine, particularly when renal function is impaired. Accumulation of normeperidine, a metabolite, is associated with CNS excitation and seizures.4z 13. Avoid the use of mixed agonist-antagonist drugs. These agents may precipitate withdrawal, and their administration complicates the transition to pure agonist agents. Pentazocine (Talwin) has been associated with a high incidence of hallucinations and confusion. 14. Some individuals cannot achieve adequate pain relief through pharmacologic means alone. When comprehensive trials of pharmacologic therapy have failed, consideration should be given to alternative modalities, including additional antitumor therapy, behavioral management, neural blockade, CNS opiate therapy, neurosurgery, and electrical stimulation.

Sublingual

Administration

The sublingual route is an attractive alternative to injections when regular oral drug administration is contraindicated. Self-administraTABLE 3. Equianalgesic Analgesics

Doses

Agent

for Commonly

Route’

Morphine

I.M. P.O. I.M. P.O. I.M. P.O.

Codeine Oxycodone Levorphanol ILevo-Dromoran) Hydrumorphone (Dllaudid) Methadone (Dolophine) Oxymorphone Meperidine

Used

Opiate

Equianalgesic Dose Im.e) 10 20-60t 130

200 15

I.M. P.O.

30 2 4

I.M. P.O. I.M.

1.5 7.5 10

P.O. I.M. Rectal I.M.

20 1.5 5

P.O.

75 300

*IM. = intramuscular; P.O. = by mouth. tSingIe-dose studies suggest an I.M.:P.O. ratio of 1:6 for morphine, but clinical experience suggests that the ratio falls to between 2: 1 and 3:~ with chronic time-contingent dosing.

Cur-r Probl

Cancer,

May/June

1989

147

tion is simple, and preliminary experience indicates that a high degree of efficacy can be expected. Buprenorphine is the only analgesic currently available for sublingual administration, although considerable research on other drugs is in progress. Preliminary results indicate that the degree of buccal and sublingual absorption correlates strongly with lipid solubility, as well as with the pK of the preparation, which determines the availability of (diffusible) unionized drug. Accordingly, morphine, the least lipid-soluble drug of those tested in preliminary studies, was most poorly absorbed, whereas fentanyl and buprenorphine were absorbed most efficiently.43

Continuous

Subcutaneous

Administration

This modality has become increasingly popular, particularly in the hands of oncologists, since, unlike intraspinal techniques, it affords control of even severe pain without the requirement for mechanical and technical assistance by another specialist. Anesthesiologists may be reluctant to rely on subcutaneous administration because of teaching that suggests that absorption is unreliable. Though this may be true for bolus administration, with adequate site care, a steady state is readily achieved when administration is continuous.M Continuous subcutaneous administration is useful when the oral route is either ineffective or inappropriate. Its advantages are simplicity, economy, and portability, and its main drawback is sedation. Family members and visiting nurses can be readily trained in maintenance. Administration is accomplished through a Z-i-gauge butterfly needle, a low-volume conduit, and a portable infusion device. Additional control can be achieved by the addition of a more sophisticated pump with “patient-controlled analgesia” features. The reader is referred to Inturrisi’s43 recent review for a thorough explanation of guidelines for administration.

ADJWANT

THERAPY

Antidepressants Classically, these agents are utilized for the treatment of depression that persists despite improved pain control. The tricyclics are also useful adjuncts for correcting anxiety and disturbances of sleep. Recently, observers have noted their potential to produce analgesia independent of psychological effects. A survey of oncologic centers in Italy was recently performed, which indicated that 63% of the 35 respondents utilized antidepressants in a total of 44% of patients, and that 98% of patients appeared to derive benefit from treatment.45 Initially, depression must be distinguished from simple grief. Grief has been characterized as a natural response to life-threatening illness and can serve as a beneficial means of coping with 10~s.~~When disturbances of mood are not self-limited and inadequately relieved 148

Cur-r Probl

Cancer,

May/June

1989

pain is present, the probability that depression is a reaction to physical symptoms should be considered. Cancer patients with pain have been shown to have a higher incidence of depression than matched cancer patients in whom pain is absent,47’48 and depression has been observed to decrease significantly in patients with cancer once pain has been relieved.4g’50 Patients’ tendency to cognitively equate increased or recurrent pain with tumor progression and impending death helps explain these findings. Since reactive depression may resolve once pain has been brought under better control, the combination of pain and mood disturbance may be a signal to initiate more aggressive pain treatment, rather than to treat depression independently. Pain is acknowledged as a complex phenomenon, and patterns of pain cannot adequately be explained in purely physiologic terms51 Just as pain influences the psychological state, as described above, psychological factors also modify pain perception.52 In some patients, loss of control and feelings of hopelessness and fear contribute to suffering and exacerbate pain. Attention to psychological needs and the successful treatment of underlying depression have been shown to reduce complaints of pain.53’54 When clinical depression persists despite supportive care and improved analgesia, treatment with an antidepressant agent is indicated. Even when anxiety is present, a tricyclic antidepressant is often a more rational choice than are the benzodiazepines. The benzodiazepines have been observed to promote further depression with chronic use, may be associated with severe mood swings, and may provoke serious symptoms of abstinence when they are withdrawn precipitously, as may occur when utilization of the oral route is no longer feasible.461 55j56 Many practitioners feel that the use of benzodiazepines should be reserved for the treatment of muscle spasm. Nightly administration of a tricyclic antidepressant in escalating doses minimizes daytime drowsiness and anticholinergic side effects while easing depression and calming anxiety. Though it is still a topic of some controversy, numerous reports on tricyclic antidepressants administered to patients with chronic pain have documented the occurrence of analgesia, independent of their psychotropic effects.57-60 It has been argued that the mechanism responsible is related to inhibition in the reuptake of serotonin and consequent enhanced activity of this neurotransmitter within the CNS. In support of a distinction between modulation of pain and affect, most clinicians utilizing amitriptyline for pain relief do so in doses inadequate to combat depression (10 to 75 mg nightly). Trazodone (Desyrel), a newer antidepressant, is an attractive alternative because anticholinergic phenomena and disturbances of cardiac conduction are rare.6 Patients are cautioned that relief of pain is often not established Cur-r Probl

Cancer,

May/June

1989

149

until 1 to 3 weeks of therapy have elapsed. Some authors recommend thorough trials of tricyclics as analgesics in any patient with chronic pain who has responded inadequately to standard pharmacologic management. Their greatest utility as analgesics has been demonstrated in patients with central pain syndromes, inand postablative anesthesia docluding postherpetic neuralgia5’ lorosa.6’ The restoration of a normal pattern of nighttime sleep is an essential step toward a return to normal functional capacity. Sleep deprivation forms part of a vicious cycle by contributing to daytime fatigue and decreased activity, permitting the patient to ruminate over pain and helplessness. Tricyclic antidepressants in low nighttime doses (i.e., amitriptyline at 10 to 100 mg or trazodone at 50 to 150 mg) may be extremely useful in correcting sleep disturbance. The common adverse effects of tricyclics, dry mouth and morning drowsiness, usually resolve over the course of 1 to 2 weeks of regular administration. Dosage should be reduced in the elderly and infirm, or in the presence of increased confusion. Other reported side effects include tachyarrhythmias, hypotension, hypertension, blurred vision, urinary retention, and constipation.

Anticonvulsants Careful trials of carbamazepine and related drugs permit control of symptoms in the majority of patients with tic douloureux, often eliminating the necessity for neurosurgical and neurolytic procedures.63 Carbamazepine, phenytoin, valproic acid, and clonazepam, alone or in combination with tricyclic antidepressants, have been used successfully in the treatment of neuralgic pain syndromes, including fifth- and ninth-cranial-nerve disorders, tabes dorsalis, and diabetic neuropathy.“4 The proposed mechanism of action involves suppression of spontaneous neuronal firing. In cancer, anticonvulsants are sometimes useful in the management of neuralgic pain associated with neural invasion by tumor, radiation fibrosis, or surgical scarring, herpes zoster, and deafferentation syndromes.14’65 Improvement can be expected in a significant proportion of patients whose predominant complaint is pain of a lancinating, burning, or hyperesthetic natum.65 Side effects of therapy are serious and can include bone marrow depression, ataxia, diplopia, hepatic dysfunction, nausea, and lymphadenopathy. Periodic hematologic monitoring is recommended. The reader is referred to a recent review for comprehensive management guidelines.65

Other Pharmacologic

Adjuncts

Antihistamines, steroids, amphetamines, and antipsychotics have been recommended as coanalgesics in selected patients with pain 160

Curt-

Probl

Cancer,

May/June

1989

related to malignant neoplasia,4’ 14,56 although clinical trials are lacking. Phenothiazines have been used in the past, but with the exception of methotrimeprazine, which has been shown to produce analgesia comparable with that of morphine,66 evidence to support their efficacy is minimal,67 and their use is discouraged because side effects can be troublesome. In contrast, hydroxyzine and dextroamphetamine have demonstrated coanalgesic properties,65’68 although their roles in clinical management are controversial. Side effects, notably constipation, nausea, vomiting, and sedation, can be anticipated with the regular administration of narcotic analgesics by any route. Preventive, regular, and aggressive treatment with antiemetics and laxatives is recommended, particularly since constipation can exacerbate pain and discomfort. ANTITUMOR

THERAPY

While it is beyond the scope of this chapter to provide a detailed explanation of the role of palliative antineoplastic therapy as a painrelieving measure, it is important that options be discussed with the primary oncologist and the patient when new symptoms develop, or before undertaking interventional analgesia. In many cases, the pathologic process and symptoms can be altered with radiotherapy, chemotherapy, hormonal treatment, and even whole-body hyperthermia.6g’ ” Partial or complete relief of bone pain has been achieved in up to 96% of patients treated with radiotherapy, which can often be administered in a single, nonfractionated dose.71s72 In a study of hormonal treatment, diethylstilbestrol provided at least temporary relief of symptoms in 75% of patients with prostatic cancer.73 Chemotherapy probably decreases pain through reduction in tumor bulk, so symptomatic improvement is generally slow. Chemotherapy is not regarded as a primary treatment for pain, although reduction in pain may occur incidentally. It is important to recognize that palliative antitumor measures have definite limitations related to efficacy, patient acceptance, side effects, and complications. The decision to pursue antitumor therapy does not imply that analgesic drugs and other supportive therapy should be discontinued. NEURAL

BLOCKADE

RECENT ADVANCES IN NEURAL PAIN MANAGEMENT

BLOCKADE

FOR

CANCER

Perineural injections of local anesthetics have been utilized therapeutically since the late 19th century following the development of the hollow needle by Rynd (1845) and Wood (1855) and the introducCurr

Pmbl

Cancer,

May/June

1989

161

tion of cocaine by Karl Koller in 1884. The application of neurolytic substances to provide relief of pain for patients with cancer has been practiced since these techniques were popularized by Dogliotti,74 Maher, and others. Neither the methods of blockade nor the useful pharmacologic agents have undergone marked change over the intervening years. Contemporary applications of these techniques represent an advance in care insofar as (1) increased experience and the availability of additional alternatives permit more sophisticated decision making with regard to screening patients and selecting the proper procedure; (2) recognition of the problem of “total pain,” and integration of anesthetic care into a multidisciplinary matrix, promise improved overall results; and (3) an increase in the dissemination of knowledge, number of skilled practitioners, and frequency of evaluations and procedures performed translates into greater benefit to greater numbers of patients. Recent enhancements of neurolytic techniques include the application of more sophisticated techniques of radiologic guidance, as in CT-assisted localization of the celiac axis. Though pituitary ablation does not involve destruction of a discrete neural pathway per se, its addition to the armamentarium of the anesthesiologist involved in cancer pain management represents a substantial advance. Finally, the epidural application of phenol is undergoing a resurgence of interest. Investigators have reported increased efficacy by utilizing repeated instillation of dilute solutions of phenol through an indwelling catheter over a period of days. Techniques of neural blockade for treating cancer pain have progressed only modestly in recent years. Increased recognition of the value of these established techniques is related to a greater availability of skilled providers and improvements in the distribution of knowledge, permitting the greatest of recent advances: enhanced opportunities to relieve pain and suffering. TYPES OF PAIN ASSOCIATED

WITH CANCER

Somatic Pain Somatic pain occurs as a result of activation of nociceptors in the cutaneous and deep tissues. The pain may be acute or chronic in nature, is typically well localized, and is often characterized as aching or gnawing. Acute pain serves to provide the organism warning of actual or impending tissue damage and is generally accompanied by hyperactivity of the sympathetic nervous system. In contrast, chronic pain persists beyond a cautionary, symptomatic role to assume the status of disease and is capable of producing destruction in its own right. Chronic pain is characterized by sympathetic adaptation and complex psychobehavioral changes. 162

Curr

Probl

Cancer,

May/June

1989

Sympathetic Pain Sympathetically mediated pain results from distention or chemical irritation of visceral end-organ structures, or from disorders of the autonomic system per se, as in reflex sympathetic dystrophy and neoplastic invasion of the sympathetic chain. In contrast to somatically mediated pain, it is described as vague in distribution and dull, burning, or electrical in nature, and it may be accompanied by vasomotor disturbances such as rubor, pallor, hyperhidrosis or anhidrosis, and hyperesthesia. Central Pain Central pain is characterized by spontaneous burning dysesthesia, hyperpathia, and hyperalgesia in the absence of peripheral tissue damage. Pain is diffuse and excruciating and may be accompanied by exaggerated skeletal muscle and autonomic responses. Patients may report symptoms consistent with alterations in sensory threshold, including anesthesia and allodynia (pain due to a stimulus that does not usually provoke pain). An additional characteristic is persistence despite standard analgesic therapy and a tendency in some instances for favorable response to tricyclic antidepressants and anticonvulsant drugs. Central pain syndromes may result from primary cortical, thalamic, and spinal cord lesions. Alternatively, the concept of central pain has been applied to describe the occurrence of symptoms in association with the disruption of neural pathways. Examples include phantom limb pain, spinal cord transection, herpetic neuralgia, and chemical or surgical neurolysis. ‘IYPES OF NERVE BLOCKS Diagnostic Neural Blockade The purpose of a diagnostic nerve block is to determine the anatomic pathway responsible for the transmission of pain, and indirectly to help establish the etiology of pain. In some cases, a block of this type will be of prognostic value, but only rarely will lasting pain relief result. It must be emphasized to the patient that the diagnostic block is only a preliminary procedure. Unless it is made clear that pain is expected to return as the local anesthetic effect dissipates, the patient’s confidence will be damaged irreparably. Diagnostic blockade is particularly useful in patients who are so debilitated or narcotized that they cannot accurately describe their pain. In cases in which pain is blunted by analgesics at the expense of alertness, it may be necessary to withhold or reverse narcotics before performing a diagnostic nerve block to permit accurate interpretation of its results. If the information obtained is to be regarded as reliable, it is critical that the block be performed accurately, and that Curr

Probl

Cancer,

May/June

1989

153

supporting signs of dermatomal numbness or sympatholysis are present. Diagnostic blockade helps distinguish among pain of sympathetic, somatic, central, and psychic origin.

Prognostic

Neural Blockade

A carefully planned prognostic nerve block with a local anesthetic agent provides essential information to the neurosurgeon, anesthesiologist, and patient regarding the expected therapeutic response to more permanent intervention. The procedure is tailored to mimic the results of neurosurgery or therapeutic (neurolytic) block, but the effect is limited to the duration of the local anesthetic that has been selected. The patient is permitted to experience an analgesic trial and to determine if the resulting numbness and other neurologic sequelae are preferable to pain. Most patients with intractable pain will respond favorably. A few patients will interpret numbness as intolerable dysesthesia or will reject the motor or proprioceptive loss that accompanies interruption of mixed nerves.

Therapeutic

Neural Blockade

For pain related to cancer, therapeutic nerve blocks generally involve the perineural application of substances that are destructive to tissue. In almost all cases a prior trial of local anesthetic blockade is mandatory to limit complications and to maximize efficacy and patient acceptance. Education of the patient and family should include frank discussion of possible advantages and disadvantages, alternatives, realistic expectations, and complications. Appropriately, this process consumes considerable time and may require repetition as disease progresses and analgesic needs change. Written consent is mandatory. Preparation should include consideration of manuevers to maximize accuracy and limit complications that may include radiologic guidance, production of paresthesia, use of a nerve stimulator, careful aspiration, local anesthetic test doses, and careful patient positioning. It is important to recognize at the outset that, even once the decision to undertake therapeutic blockade is undertaken, the end result may be inadequate. The dying patient’s confidence in physicians is fragile, so prior understanding that therapeutic procedures often require repetition is essential to avoid disappointment.

Practical Application Diagnostic Block.-An example of the appropriate use of a diagnostic nerve block is in the patient presenting with radicular thoracic pain and metastases to ribs and vertebrae. If local anesthetic intercostal block of the involved rib and its neighbors does not eliminate the patient’s symptoms, then the pain is inferred to originate 164

Curr

Probl

Cancer,

May/June

1989

from the more proximal vertebral ture. A sympathetic etiology can tions of the sympathetic chain. somatic blocks or subarachnoid taken to verify that pain is related

lesion or to be sympathetic in nabe tested with paravertebral injecIf pain is unrelieved, paravertebral or epidural injections are underto vertebral collapse.

Prognostic Block.-If one of the diagnostic procedures has been observed to have eliminated the pain under investigation, then that information can be interpreted as having prognostic value. A separate prognostic block is unnecessary if the patient has been oriented to appraise the results of the initial procedures critically. Therapeutic Block.-Depending on which investigative procedure has provided effective pain relief consistent with patient acceptance, a therapeutic block is planned. Guidelines for therapeutic blocks are outlined in the text. LOCAL ANESTHETIC

BLOCKADE

As noted above, the greatest utility of local anesthetic injections in patients with cancer pain is for diagnostic and prognostic purposes, particularly when life expectancy is limited. Local anesthetic blocks may play a major role in the treatment of patients whose pain is related to anticancer therapy, or in cancer patients with coexisting pain of nonmalignant etiology. It is well accepted that persistent pain relief often follows repeated local anesthetic sympatholysis in patients with reflex sympathetic dystrophy (RSD) of nonmalignant etiology. Recently, with identical treatment, pain relief of extended duration has also been observed in a high percentage of cases of tumor-induced RSD.76 The relatively benign nature of this form of treatment warrants a trial series of local anesthetic blockade of the affected sympathetic chain when autonomic dysfunction is suspected. Also, reflex myofascial pain and spasm probably occur with more frequency than has been previously recognized and are amenable to trigger-point injections with long-acting local anesthetics. Finally, continuous epidural infusions of local anesthetic alone77 or in combination with narcotics7’ have been successfully used to maintain persistent blockade of noxious afferent impulses in the home treatment of dying patients with severe intractable pain. NEUROLYTIC

BLOCKADE

General Considerations Despite limited revision in the principles and technique of neurulytic blocks, guidelines for their use in clinical practice are presented to define their role in the context of comprehensive cancer pain management and in virtue of their unique utility to the anesthesiolCurr

Probl

Cancer,

May/June

1989

165

ogist involved in managing cancer pain. Standard sources are recommended for the practitioner interested in technical details of individual procedures.17’ 7g Consideration for neurolytic intervention in patients with cancerrelated pain is indicated when symptoms cannot be controlled adequately with medications and other conservative therapy, and when life expectancy is limited. Failed conservative care exists when pain persists despite escalation of therapeutic measures or when side effects related to those measures are viewed by the patient, family, or physician as being intolerable. Restrictions related to life expectancy are not rigid. Upper limits of 6 to 18 months have been recommended, but each case should be considered on its individual merit. At the opposite end of the spectrum, procedures may be performed up through the last days of life, particularly at the bedside, if warranted by unrelieved suffering. Regardless, referring physicians should be encouraged to seek consultation early in the course of terminal disease since many techniques require patient cooperation to produce optimal results. Also, some series indicate a higher incidence of failure when pain has persisted in excess of 3 to 4 months.” Operant factors may include psychological maladaptation and isolation of targeted neural tissue by radiation fibrosis or malignant cells. Neurolytic Drugs Alcohol is commercially available in single-milliliter ampules as a colorless solution that can be injected readily through small-bore needles, and that is hypobaric with respect to cerebrospinal fluid. Depending on the site of injection and the concentration of alcohol, administration is accompanied by a variable degree of discomfort that, at its extreme, is excruciating but transient. It is generally used undiluted (absolute or 100% alcohol) and, if left exposed to the atmosphere, will be diluted by absorbed moisture. Denervation and pain relief sometimes accrue over a few days following injection. Injectable phenol requires preparation by a pharmacist, the process for which has been described by Rajsl in a popular text. Various concentrations of phenol prepared with saline, water, glycerine and different radiologic dyes have been advocated. Phenol is relatively insoluble in water, and at room temperature, concentrations in excess of 6.7% cannot be obtained without the addition of glycerine. Phenol mixed in glycerine is hyperbaric with respect to cerebrospinal fluid but is so viscous that, even when warmed, injection is difficult through needles smaller than ZO-gauge. Shelf life is said to exceed 1 year when preparations are refrigerated and are not exposed to light. Clinically, a biphasic action has been observed, characterized by an initial local anesthetic effect producing subjective warmth and numbness, which gives way to chronic denervation. Quality and ex166

Cur-r

Probl

Cancer,

May/June

1989

tent of analgesia may fade slightly within the first 24 hours of administration. Less commonly, ammonium sulfate and chlorocresol are utilized to produce neurolysis. The pathologic changes that follow the administration of neurolytic drugs, and proposed mechanisms of neurolysis, have been described in the literature.75’SZ’83

Peripheral

Blockade

Peripheral neurolysis has a definite, although limited, role in the management of pain of malignant origin.@ To ensure effective analgesia, neural interruption is planned proximal to the source of irritation. In technically difficult cases, such as blockade of upper intercostal nerves (where overlying scapula and muscle increase the risk of pneumothorax), more proximal paravertebral or subarachnoid blockade should be elected. Anatomic landmarks tend to be more obvious in the presence of cachexia and weight loss. Because the sensory distribution of peripheral nerves overlaps, blockade of neighboring segments is recommended. Many peripheral nerves are of mixed function. A pretherapeutic prognostic block with local anesthetic is essential to evaluate the impact of concomitant motor deficit. In performing a peripheral neurolytic block, accuracy is essential for good results and to avoid damage to nontargeted structures. This is particularly true in the cervicofacial region, where abundant neural and vascular structures are closely spaced, and when alcohol or phenol are used because their diffusibility in biologic tissue is less than that of the local anesthetics. Regeneration of peripheral nerves is sometimes accompanied by the development of neuritis or neuroma formation. It has been suggested that alcoholic neuritis is related to incomplete destruction of somatic nerves, and that incidence is less when a solid, prolonged block has been obtained.17 Alcohol seems to have more propensity to produce local irritation than does phenol, and local reaction seems to be less common in cranial nerves than in other peripheral nerves. Central nervous system maladaptation to deafferentation, as well as local phenomena, may result in burning pain that has a potential to be even more objectionable than the original complaint of pain. The threat of postablative dysesthesia is of limited consequence when life expectancy has definite limits or intractable pain already exceeds tolerance levels.

Cranial Nerve Block Pain related to malignant neoplasm of the head and neck poses one of the most challenging management problems to all providers of care. Conventional analgesic therapy may prove inadequate, in part because of failure of physiologic splinting, a normal coping strategy unavailable when pain is aggravated by relatively involuntary activities such as swallowing, eating, talking, and moving the Curr

Probl

Cancer,

May/June

1989

157

head. Because of overlapping sensory innervation (cranial nerves V, VII, IX, and X, dorsal roots of the second and third cervical nerves), cephalic pain rarely is limited to the distribution of a single nerve. Major surgical intervention introduces considerable risks of mortality and morbidity and is regarded by many neurosurgeons as a last resort. Surgery is often accompanied by prolonged hospitalization, additional functional deficit, and disfigurement, the prospects of which are poorly tolerated in the preterminal patient. When pain is intractable, nonsurgical treatment (radiotherapy, systemic and regional chemotherapy) should be aggressively pursued. If conservative measures have been exhausted, consideration should be given to analgesic methods that produce generalized effects (epidural or intraventricular narcotics, deep brain stimulation). When pain is related to metastatic instability of the cervical spine, anterior stabilization should be considered. In a retrospective review of surgical decompression and stabilization for patients with cervical spine metastases, significant pain relief in excess of 3 months was achieved in 11 of 17 patients, and 5 of 11 paralyzed patients experienced significant neurologic recovery.85 In selected patients, blockade of the involved cranial and/or upper cervical nerves is of great value. When ablative blocks or surgery are being considered, local anesthetic blocks help to ascertain the relative contribution of individual nerves to the painful state. Lasting pain relief is difficult to achieve with discrete nerve blocks because of sensory overlap. Also, technical difficulties may be encountered because of the tendency for tumors in these regions to erode and distort surrounding tissue, and owing to the high proportion of patients that come to treatment after mutilating surgery or radiotherapy. Oncogenic pain limited to the distribution of the trigeminal nerve or one of its branches may be amenable to thermal or chemical interruption of the involved nerve. When feasible, thermocoagulation may be the treatment of choice because analgesia can frequently be obtained without sensory loss. A recent analysis of a series of 21 patients treated in this fashion reported no morbidity or mortality and “very good” long-term results in over half of patients.86 When chemical block of a branch of the fifth nerve is to be undertaken, consideration should be given to the likelihood of further tumor extension and prophylactically extending the field of analgesia by blocking the gasserian ganglion.” Keratitis and corneal ulcer are possible complications but are of limited consequence. If extension of analgesia is necessary, consideration may be given to blocking the second and third cervical roots and/or the ninth and tenth cranial nerves. Bilateral block of cranial nerves IX and X should not be attempted because resulting paralysis of pharyngeal and laryngeal muscles impairs swallowing and phonation. When neural blockade is planned to eliminate pain over the oc168

Curr

Pmbl

Cancer,

May/June

1989

ciput and neck and there has been extensive paravertebral approach to the cervical nerves and the epidural route should be considered.

radiotherapy, the may be difficult,

Sacral Nerve Block Selective blockade of the sacral nerve roots via their dorsal foramina is a useful alternative to subarachnoid or caudal block in patients with pain from pelvic or rectal malignant neoplasm in whom bowel and bladder function are still intact. Pain may be due to direct invasion, chronic infection, fistula formation, or postradiation cystitis. Diagnostic blocks of the individual sacral roots with local anesthetics performed in succession will determine the pathways involved in the transmission of pain. The second and third sacral roots are most commonly implicated in maintenance of bladder function, and unilateral predominance of innervation of detrusor reflex activity has been observed with some frequency,8’ so treatment that avoids bilateral S-2 and S-3 blocks is likely to preserve function. Of a group of 15 patients with bladder pain, treated first with a series of prognostic local anesthetic blocks, ten patients were selected to receive discrete sacral nerve root blocks with 6% aqueous phenol. Seven of the ten patients experienced long-term pain relief (average, 26.5 months), and no mortality or significant morbidity was observed.= Good results without complications were obtained in a small series of cancer patients with perineal pain treated with injections of phenol through the fourth sacral foramen.8s

Subarachnoid

and Epidural

Neurolytic

Block

Classic publications describe the techniques of these and other nerve block procedures in detail,17’g0’ 91 and the reader is encouraged to consult them to review technical aspects. The advantages of neuroaxial neurolysis are (1) a high proportion of good results in properly selected cases; (2) ease of performance with minimal requirements for equipment; (3) minimal or no requirements for hospitalization; (4) duration of pain relief that is generally adequate for the preterminal state; (5) ease of repetition when necessary; (6) suitability for aged or debilitated patients; and (7) a low complication rate when proper technique is observed. Lytic neuroaxial block produces pain relief by chemical rhizotomy. Despite early speculation that phenol was capable of exercising selective blockade of small sensory fibers,‘5 pathologic studies have demonstrated that, regardless of size, nervous fibers are affected indiscriminately by both alcohol and phenol.“’ ” The degree and extent of sensory loss depends on the actual number of fibers destroyed rather than fiber type, which is, in turn, determined by the concentration and quantity of the neurolytic agent.83 This relationship is supported by recent reports of higher success rates in paCurr

Probl

Cancer,

May/June

1989

169

tients with pelvic malignant neoplasms who were treated with 10% and 15% subarachnoid phenol vs. a 7.5% preparation,g3 and in patients with a variety of neoplasms treated with 15% vs. 10% subarachnoid phenolW Controlled studies comparing the results of alcohol and phenol neurolysis are not available. Most authorities agree that neither agent offers a clear advantage, except insofar as variations in baricity facilitate positioning of the patient in selected cases.gl’ s5 Dozens of reports of large series of patients treated with chemical rhizolysis have appeared in the medical literature. Results are difficult to compare because of variations in patient selection, extent and type of underlying malignant neoplasms, injection techniques, and criteria for success. Swerdloti6 has analyzed the results of 13 published series documenting the treatment of over 2,500 patients. In 58% of patients, “good” relief was obtained, “fair” pain relief was achieved in an additional 2l%, and in 20% of patients “little or no” relief was noted. Average duration of relief is difficult to estimate but generally is regarded by authorities to be 3 to 4 months,17 with a wide range of distribution. Reports of analgesia persisting in excess of 1 or even 2 years are not infrequentg6 Chemical rhizolysis can be performed at any level up to the midcervical region, above which the spread of caustic agent to medullary centers carries significant risk of cardiorespiratory collapse.s7 Subarachnoid injections of phenol at the C3-4 interspace have been carried out without complications.80 Access to the subarachnoid and epidural spaces above the lumbar region presents a technical challenge that decreases as experience accumulates. Cervical Block.-Some investigators have reported only fair results when employing subarachnoid phenol rhizolysis for brachial and upper-thoracic pain,” a phenomenon that has also been observed after neurosurgical procedures for brachiothoracic pain9” loo The aggressive growth characteristics of lesions involving the brachial plexus and chest wall may play a role. Nevertheless, in carefully selected patients, cervicothoracic subarachnoid and epidural neurolysis have been utilized for relief of upper-extremity pain with relative success.75’so These options are not considered unless pain is intractable and so severe as to render the involved limb immobile and useless. An early trial of stellate ganglion blocks is useful to exclude sympathetically mediated pain that could be permanently blocked without sacrificing function.‘0”‘02 When pain does not involve the whole of the extremity, paravertebral or more distal nerve blocks may be considered to preserve maximum function. Brachial plexus block with dilute phenol has been reported in a small number of patients as being moderately effective but introduces the risk of spread to adjacent structures and may require frequent repeti160

Curr

Probl

Cancer,

May/June

1989

tion.lo3 When pain is related to infiltration of the brachial plexus with tumor or radiation fibrosis, neuroaxial block with phenol or alcohol will often provide effective relief. Motor and proprioceptive deficit are anticipated, thus emphasizing the importance of careful patient selection. Thoracic Ho&.-Patients with radicular thoracic or upper-abdominal pain are ideal candidates for discrete chemical rhizotomy because medication can be introduced distant from the major limb plexuses and the origin of nerves subserving bladder and bowel function. Intercostal muscle paresis may occur despite meticulous positioning but seldom produces increased respiratory compromise. Indeed, subjective and objective reductions in breathlessness are often observed once splinting has been eliminated by successful neural blockade. Lumbar Ho&.-When employing subarachnoid techniques for pain in regions subserved by the lumbar nerves, it is of particular importance to observe the rule that instillation of a neurolytic substance should be planned at the level at which the involved nerve roots exit from the spinal cord rather than their site of egress from the vertebral column. The more proximal site of injection affords greater surface area for exposure of the targeted neural tissue to the neurolytic drug;17 more importantly, since the terminal portions of distal nerves are less exposed to the neurolytic agent, more discrete analgesia can be expected, and the incidence of complications should be reduced. Charts indicating the relationship between the spinal column and the nerve roots as they emerge from the spinal cordI should be carefully checked before each procedure. Utilizing this method, access to lumbar nerve roots is gained through dural puncture between the low thoracic vertebrae. Additional precautions undertaken to maximize efficacy and limit complications include the use of an operating room-type table that can be “broken” and readily manipulated, careful attention to maintenance of patient posture, and conservative selection of dosage and rate of injection. Maximum suggested doses should be respected,*7”04 and communication with the patient is essential to verify the desired location and extent of blockade. From the standpoint of safety, it is preferable to expose the patient to a repeated procedure or to inject small volumes through multiple needles than to exceed standard recommendations .17,lo4 Patients need to be cautioned in advance that though the actual puncture and injection ordinarily produce only minimal discomfort, maintenance of optimal posture for 15 to 30 minutes after the neurolytic has been instilled may be distressing. Bowel and bladder dysfunction are among the most feared complications of neuroaxial lytic block, although actual incidence and Curr

Probe

Cancer,

May/June

1989

161

severity are low when proper technique is observed, regardless of the agent utilized. Results of two large series are given in Table 4.

Saddle Block.-Perineal and pelvic pain are amenable to neuroaxial lytic blockade by a number of routes. A useful approach for unilateral pain involves placing the patient in a lateral position (dependency dictated by choice of agent) and injecting the neurolytic agent through a low lumbar puncture. If pain is bilateral, the procedure can be repeated after a few days, with the patient’s position reversed. It is not uncommon for bilateral pain to resolve significantly after a unilateral block, making a second procedure unnecessaxy?os Alternatively, a true saddle block can be utilized for pain that crosses the midline.‘l Hyperbaric phenol is introduced through a low lumbar puncture, and the patient is maintained in the sitting position modified by a 4.5’ posterior tilt for 15 to 30 minutes. Finally, the injection of small volumes of dilute phenol via the caudal route has recently been described in a series of 26 patients with perianal pain due to malignant neoplasm.1o6 Pain relief persisted for a mean of just 12.7 days, but only one complication (transient urinary retention) was described. Since, despite all precautions, at least temporary urinary or even anal dysfunction can be anticipated in a high proportion of patients, these techniques should be reserved for patients with indwelling catheters or patients who are bedridden and accept the risk of additional deficit, Subarachnoid vs. Epidural iVeuroZysis.-Subarachnoid neurolysis offers the following potential advantages over classic epidural techniques. (1) Return of cerebrospinal fluid verifies subarachnoid needle placement, whereas localization of the epidural space must be inferred from the results of epidurograms and/or test doses of local anesthetic. (2) Subarachnoid neurolysis generally results in more profound analgesia, and, as a consequence, reinjection is required TABLE 4. Incidence of Bowel Studies of Intrathecal

No. of patients No. of blocks Agent Diagnosis Dysfunction Bowel Bladder Transient Permanent 162

and Bladder Neurolvsis

Dysfunction

Hav’”

Stovner

in Two and Endmsen”

2.52 407 Alcohol Cancer

151 313 Phenol Cancer

0

0

2 1

6 0 Curr

Probl

Cancer,

May/June

1989

less often. This phenomenon is probably related to increased direct contact between drug and targeted nerve roots. (3) Subarachnoid injection is readily performed on an outpatient basis, or even at the bedside. Recent recommendations that epidural neurolysis be accomplished by repeated administration of phenol through an indwelling catheter mandate inpatient hospitalization. (4) Although reports indicate that gravity and position can be partially relied on to control the effect of epidural block with hyperbaric phenol,‘07 these factors can be utilized to exert more precise control in the case of subarachnoid injection. (5) The excessive viscosity of pure phenolglycerine preparations prevents injection through small-caliber tubing. If epidural block is planned with the intention of reapplication through a catheter, then the phenol-glycerine mixture must be diluted with water, saline, or dye, reducing baricity and introducing the potential for reduced control of spread. A newly designed epidural catheter made from spiral stainless steel coils coated with fluoropolymers has recently been introduced with the intention of facilitating radiologic localization, aspiration, and repositioning.1o8 No controlled studies have been conducted comparing epidural and subarachnoid neurolysis. The main advantage of phenol epiduml neurolysis is its applicability for pain that occupies a wide distribution or is bilateral. Despite the considerations noted above, the main disadvantages cited for classic epidural neurolysis are impressions of a shorter duration of action and inferior intensity of analgesia when compared with subarachnoid techniques.g6 Recent studies suggest that these shortcomings may be overcome by repeated administration over time through an indwelling catheter.*Os-lll Epidural injection has been postulated to limit spread to dorsal nerve roOtS,110 which should reduce the likelihood of motor dysfunction. Using an in vitro model, Racz et al?l’ have demonstrated that sections of canine dura apposed to reservoirs of cerebrospinal fluid are relatively impermeable to 2.75% and 5.5% phenol in saline. Risks of meningeal irritation, postdural puncture headache, and intracranial spread are less with epidural than subarachnoid neuro1ysis.g6 The most commonly recommended neurolytic agent for epidural use is still phenol. Despite reports of favorable results utilizing alcoho1,112 its instillation in awake patients produces agonizing paing7 unless preceded by injections of local anesthetic, which reduce predictability. Postinjection neuropathy and neuralgia have been dbserved after epidural alcohol.g6

Sympathetic

Nerve Block

Local anesthetic infiltration of the sympathetic nervous outflow can be performed for diagnostic, prognostic, or therapeutic purposes. A diagnostic nerve block helps to establish the relative contribution of the autonomic vs. somatic nervous system to pain transCurr

Pmbl

Cancer,

May/June

1989

163

mission. Response to local anesthetic blockade helps to determine whether repeated local anesthetic blocks, a neurolytic block, or surgery is likely to provide prolonged relief. Finally, in carefully selected patients, a therapeutic effect may be obtained either by serial injections of a local anesthetic drug or by the injection of a neurolytic agent.

Cervicothoracic (Stellate) Ganglion Block.-Repeated sympathetic nerve blocks with local anesthetic are a well-documented method of relieving the pain of reflex sympathetic dystrophy113”14 and may also be of value for tumor-induced sympathetic-mediated pain as well (see “Local Anesthetic Blockade” section). This phenomenon has been most thoroughly documented for brachiocephalic pain because carefully performed local anesthetic stellate ganglion blocks are well tolerated, and practitioners are reluctant to inject neurolytic agents in the vicinity of the stellate ganglion. Neurolytic stellate ganglion block is regarded as hazardous because the cervicothoracic ganglion may be difficult to locate with precision, and spread to nontargeted structures may produce severe complications because of the close proximity of major neurovascular structures. Specific risks include erosion, thrombosis or spasm of major vessels, cerebral infarction from intravascular injection, prolonged hoarseness from spread to the recurrent laryngeal nerve, upper-limb dysfunction if elements of the brachial plexus are affected, and sloughing of the superficial tissues. Nevertheless, there have been anecdotal reports of stellate gangliolysis performed with up to 10 ml of 6% aqueous phenol with good results and no complications.‘01 Bonica17 reserves neurolytic stellate block for exceptional cases and then limits the injectate to either a small volume of phenol or no more than 1.5 ml of absolute alcohol, and Lijfstrijm and Cousins115 recommend 1 to 2 ml of 6% aqueous phenol or 10% phenol in iothalamate meglumine (Conray) dye. If surgical sympathectomy is contraindicated, and neurolytic stellate block is to be undertaken, the following precautions should be observed: a thorough trial of local anesthetic blockade, careful explanation of the procedure and its possible sequelae, radiologic visualization of a characteristic spread of contrast medium,‘16 meticulous aspiration and needle immobilization, evidence of sympathetic block following the injection of 1 to 2 ml of local anesthetic solution, injection of 1 to 2 ml of dilute neurolytic preparation (3% to 6% phenol), and flushing the needle before its removal. Lumbar Sympathetic Block.-Neurolysis of the lumbar sympathetic ganglia may be undertaken providing trials of local anesthetic injections have been shown to provide pain relief. In the presence of malignant disease, the most common indication for lumbar sympatholysis is pelvic pain of urologic, gynecologic, or rectal origin. 164

Curr

Probl

Cancer,

May/June

1989

Other indications include lower-extremity pain from lymphedema or reflex sympathetic imbalance. With simple attention to detail, risks of interference with bladder and bowel function are virtually nil. Classically, lumbar sympathetic block, as described by Mandl in 1926 and subsequently modified,17 is accomplished by injection through three needles positioned with their tips anterior to the psoas muscle, at the anterolateral aspect of the bodies of the second, third, and fourth lumbar vertebrae. Consistent with Winnie’s117 concept of “plexus anesthesia,” it has become common practice to substitute a single injection of a large volume of local anesthetic solution (up to 20 to 30 ml) through a single needle positioned in the correct fascial plane near the second lumbar vertebra.*18 Results are comparable, and patient discomfort is reduced.l15 A recently published fluoroscopic and pathologic analysis of the relevant regional anatomy in cadavers suggests useful guidelines for the single-needle technique.ll’ In a small series of patients, complications were not observed after injections of 10 ml of 10% phenol through a single needle.“’ Concern that large volumes of injectate may spread outside the correct fascial compartment to involve lumbar somatic nerves still influences most practitioners to rely on two or three needles when a neurolytic agent is employed.‘15 Celiac Plexus Block.-Neurolytic celiac plexus block (NCPB) has received widespread attention because of its excellent potential to relieve upper-abdominal and referred back pain secondary to malignant neoplasm involving structures derived from the foregut. The most common indication for celiac axis block is pancreatic cancer, which, contrary to traditional teaching, is frequently associated with painful rather than painless jaundice. Also, NCPB is efficacious for pain associated with neoplasms involving the distal esophagus, stomach, liver and bile ducts, small bowel, proximal colon, adrenals, and kidneys. The posterior percutaneous approach, introduced by Kappis (1914) and popularized by MOOIX? and others, using two needles (6 in., 20-gauge) is most commonly advocated (Fig 2). There has been renewed interest in utilizing an anterior approach, as radiologists have become more experienced with this route for biopsy and drainage procedures. Ischia et a1.l” have reported on a transaortic technique that is similar to conventional approaches, except that the needle is deliberately passed through the aorta, in a manner resembling the transarterial method of brachial plexus blockade. No complications occurred in 28 patients treated with this method. Pain relief was obtained in 93% of patients, and favorable patterns of dye dispersion were consistently observed. Another recent modification, described by Singler,lz3 involves deliberate perforation of both diaphragmatic crura under CT guidance to ensure the spread of injectate anterior to the aorta. An underutilized approach is injection under direct vision by the surgeon at the time of laparotomy. At times Curr

Probl

Cancer,

May/June

1989

165

FIG 2. A and B, needle placement for celiac plexus block illustrating anterior to T12 vertebral body. (From Ferrer-Brechner T: Anesthetic cer pain. Semis Oncol 1985; 12:431-437. Used by permission.)

location of needle tips management of can-

this may not be possible because diagnosis is made on a nonsurgical basis or because of the presence of diffuse intra-abdominal disease. Injecting intraoperatively, Flanigan and KIY#‘~ reported an 88% incidence of postoperative pain relief with a mean duration of 4.3 months and recommended that, when intraoperative injection is not possible, a radiolucent clip be placed in the region of the celiac axis to facilitate postoperative percutaneous localization, Surgeons need to be made aware of the potential of this manuever for the relief of postoperative suffering. An 85% to 34% incidence of good to excellent pain relief has been obtained in several large series of patients undergoing NCPB for pain from pancreatic cancer per selz5 or a variety of intra-abdominal conditions,““’ lz7 although repetition is required in some patients. In a series of 136 patients,lz5 analgesia was present until the time of death in 75% of cases, and in an additional 12.5% pain relief was maintained for more than 50% of survival time. The location of the celiac axis deep within the retroperitoneum near the vertebral column and in close proximity to major vessels (aorta, vena cava, and their branches) and viscera (kidneys, pleura) provides the potential for complications of devastating proportions. Reported complications include pneumothorax, chylothorax, pleural effusion, convulsions, and paraplegia.125-‘31 Nevertheless, the results of several large series indicate that, given sufficient attention to detail, the incidence of complications should be minimal. In a series of 136 patients, the only complication of any significance was pneu166

Cur-r Probl

Cancer,

May/June

1989

mothorax, which occurred in two patients, neither of whom required tube thoracostomy.‘25 Also, of 114 blocks performed in another group of 100 patients, the only significant complication was partial unilateral lower-extremity paralysis in an obese patient who could not be positioned properly.‘26 Postural hypotension and diarrhea occur frequently but are usually self-limited. Moore et al.13’ have studied needle trajectories and the distribution of injectate in patients, using roentgenographic and CT controls, and at autopsy in cadaveric specimens. They concluded that at the level of the inferior border of the 12th rib, an insertion site no greater than 7.0 to 7.5 cm lateral to the corresponding spinous process is desirable to avoid renal puncture. Their findings also suggested that to produce optimal results, (1) two needles should be utilized, (2) 25 ml of solution should be injected through each needle, and (3) proper depth of insertion may slightly exceed that which has been traditionally taught. The findings of another recent cadaver study’33 emphasize anatomic variability but suggest that superior results can be obtained by placing the left-sided needle tip at the junction of the lower and middle thirds of the first lumbar vertebral body, and the right-sided needle tip 1 cm higher. The authors also suggest that when the celiac artery can be identified, injection should be planned 0.5 to 1.0 cm below its origin. The role of roentgenographic, fluoroscopic, and CT guidance is controversial. Most authors agree that when neurolytic solutions are to be injected, some form of radiographic verification is desirable from medicolegal and technical standpoints, although the possibility of a complication is still not excluded. The use of CT adds to cost, procedure time, and logistic difficulties but provides improved documentation of the spread of injectate and relationships between needle trajectory, viscera, and vasculature. No controlled studies have compared the results of NCPB performed with and without different radiographic controls; nevertheless, our opinion is that some form of radiologic guidance is compulsory, and the utilization of CT guidance is highly desirable. CNS OPIATE SPINAL

THERAPY

OPIATES

Research into endogenous opiatelike substances and the discovery of CNS opiate receptors has led rapidly to the administration of exogenous opioids directly to CNS sites to provide clinical pain relief. The hallmark of spinal narcotic therapy is analgesia without motor, sensory, or sympathetic blockade. A gamut of opiate analgesics have been administered intradurally and extradurally, but, as with traditional forms of opiate administraCurr

Probl

Cancer,

May/June

1989

167

tion, morphine has emerged as the standard to which other drugs are compared. Morphine is preferred because of its long duration of action (about 24 hours) and the availability of a preservative-free preparation approved by the FDA for intraspinal use. Morphine’s extended duration of pharmacologic activity is explained by its low lipid solubility relative to other narcotics, which results in delayed uptake from intradural and extradural depots, and also greater cephalad migration. The density of opiate receptors within the spinal cord is highest in the marginal zone and substantia gelatinosa, although some pharmacokinetic studies suggest that intraventricular receptors, rather than spinal loci, may be morphine’s main site of action.‘34 In one study, samples of cerebrospinal fluid obtained from the Cl-2 interspace contained high levels of morphine as early as 1 hour after lumbar injection.135 Rostral spread also seems to correlate with the occurrence of side effects. Complications associated with neuroaxially administered opiates include pruritus, nausea and vomiting, urinary retention, and biphasic respiratory depression. Delayed respiratory depression is the main impediment to more liberal use of epidural narcotics in acute-care settings for labor and postsurgical pain, but problematic respiratory depression is extremely uncommon in the narcotic-tolerant patient with chronic pain .134A multiplicity of distinct classes of receptors for opiates and opiatelike substances have been identified and are postulated to selectively mediate the therapeutic effects and the side effects of the endogenous and exogenous opiates. It is expected that, in the future, the isolation of highly receptor-specific pharmaceutical substances will permit complication-free drug administration. The reader is referred to a recent comprehensive review for additional information.‘36 Limited trials with intrathecally administered synthetic j3-endorphin, the most potent of the known endogenous opioids, suggest that long durations of high-quality pain relief (up to 4 days) can be obtained with minimal side effects.137”38 The field of CNS opiate therapy is still so new that guidelines for administration and selection of route, drug, and protocol are still emerging. A lumbar epidural catheter is the most accepted means of access to the CNS, particularly among anesthesiologists, because of the impression that complications are fewer than with subarachnoid administration and because intrathecal administration requires neurosurgical implantation. Once epidural catheter placement is verified by observation of the response to local anesthetic administration, a standard catheter is secured with sterile dressings. It can be used in this fashion for a variable period while efficacy, patient acceptance, and a dose-response relationship are evaluated. Under local tepidural) anesthesia, a Silastic catheter can eventually be tunneled subcutaneously to an exit site in the anterior abdominal wall (Fig 3). Using sterile technique, morphine can be administered daily 168

Curr

Pro61

Cancer,

May/June

1989

FIG 3. Percutaneous Siliastic catheter for chronic administration of opiates into the epidural space. (From Poletti CE: Open cordotomy: New techniques, in Schmider MM, Sweet WH (eds): Operative Neurosurgical Techniques: Indications, Methods and Results. New York, Grune & Stratton, 1982, vol 2. Used by permission.)

or twice daily, as needed. The gradual development of tolerance is common, but analgesia can usually be maintained by increasing dosage. Occasionally, massive tolerance occurs, resulting in daily morphine requirements in excess of 100 mg.13’ Attempts to reverse tolerance have been only partially successful and have included the administration of intradural and extradural clonidine,14’ conversion to the intraventricular route,14’ and temporary conversion to local anesthetic infusion to facilitate restoration of receptor sensitivity. In the bedridden patient, an infusion pump can be utilized for the administration of morphine, more potent short-acting narcotics, or Curr

Probl

Cancer,

May/June

1989

168

even combinations of a dilute local anesthetic and narcotic.78 An alternative technique involves intrathecal placement of the catheter tip and drug administration through a freon-driven pump implanted in the subcutaneous tissue of the abdominal wall (Fig 41. A 50-ml reservoir is filled percutaneously by a physician every 14 to 21 days, and a constant volume of drug (2 to 4 ml/day) is infused continuously. Alterations in dosage are accomplished by replenishing the reservoir with drug of the appropriate concentration, or bolus administration into a separate port. Coombs et al.14’ reported good analgesia in a group of cancer patients, initially administering a mean of 2.0 mg of morphine daily and 6.6 mg daily at the end of 12 weeks. Investigators are perfecting new systems that incorporate a miniaturized computer that can be programmed externally to alter infusion rates. Expected equipment costs, independent of surgical fees, are estimated to range between $6,000 and $8,000.‘43 Though subarachnoid infusion techniques offer advantages to the patient because the apparatus is completely internalized, they are costly and involve additional surgery. The potential for leakage of cerebrospinal fluid, chronic headache, meningitis, and equipment malfunction are greater than when externalized epidural catheters are used.13’ The main indication for chronic administration of intraspinal opiates is intractable pain in the lower body, particularly when symptoms are bilateral or cross the midline. With proper screening, good to excellent pain relief can be expected in up to 90% of patients.lM Anecdotal reports indicate that, in selected cases, analgesia may be obtained in the presence of more cephalic loci of pain. Because tolINLET

OUTLET

SEPTUM

CATHETER CHARGING

FLUID CHAMBER

FIG 4. lnfusaid Model 400: a bellows pump system 3 cm thick 160 gm, and having a capacity of 47 ml. Infusion rates rates depend on body temperature as well as atmospheric 170

and vary

Curr

9 cm in diameter, weighing from 0.5 to 0.6 ml/day; flow pressure. Probl

Cheer,

May/June

1989

erance eventually develops,‘45 the chronic administration of spinal opiates is most appropriate for patients with a predicted life expectancy of less than 3 to 6 months. Bolus administration of morphine is thought to induce tolerance more quickly than when the drug is administered by continuous infusion.145 The safety of chronic outpatient catheter use has been established and depends primarily on training nursing personnel or family members in catheter care and in recognizing problems. The use of a bacterial filter is an essential safeguard. The routine administration of a local anesthetic test dose to detect subarachnoid migration of the epidural catheter tip is controversial, particularly with the use of large bone catheters. INTRAVENTRICULAR

OPDITES

Limited experience with small doses of morphine administered through Ommaya reservoirs suggests that this is an effective and practical method of relieving pain in selected patients. Access to the ventricular system is usually through a coronal burr hole made using local anesthesia, although a new technique involving passage of a percutaneous catheter through a 14-gauge Touhy needle into the cisterna magna has been described.‘46 The goal of most practitioners is ongoing outpatient administration by a family member or visiting nurse.141,147-149

The quality of analgesia is excellent in most cases and is apparently unaifected by the site of pain. In one study, patients who had previously been treated with lumbar injections of morphine consistently rated intraventricular administration higher with respect to quality of pain relief.15’ Various criteria for the selection of patients have been used. A trial of lumbar or cisternal morphine injection usually precedes implantation.141’ 14’ Many patients are referred for chronic ventricular administration after intraspinal administration has been abandoned because of tolerance or local problems. The majority of implants are for intractable cervicofacial pain because more conservative or traditional techniques frequently prove inadequate. Other candidates include patients with cervicobrachial pain, bilateral or midline pain, or unremitting pain in any body part that persists despite neurosurgery or that is not amenable to surgery. Optimum life expectancy is on the order of about 6 months. Premature initiation increases risks of sepsis, tolerance, and respiratory depression. In imminently preterminal patients, tolerance may already exist, which, together with a higher incidence of mental confusion, complicates management.141,148

Various hardware, drug preparations, and dosing schedules have been devised. In one study, 0.33 to 4.0 mg of morphine produced analgesia for periods ranging from 36 to 150 hours,151 but generally Curr

Probl

Cancer,

May/June

1989

171

at least daily administration is required.148’152 In another study, using a very dilute morphine preparation, three daily doses totaling 200 l.~g was effective.“’ Other investigators have routinely administered bolus doses of 1.0 to 2.0 mg daily, with occasional patients requiring 7.0 mg or more.1e150 Dose requirements appear to be related to the amount of narcotic taken orally or systemically before treatment.14’ In a study comparing the effects of epidural and intraventricular morphine, nausea, vomiting, and pruritus occurred with about the same frequency in both groups. Urinary retention was more specific for spinal administration, and dizziness, sedation, and dysphoria occurred more often with ventricular administration.147”4s Respiratory depression occurs occasionally and appears to be dose related, one case having resulted from an accidental overdose.147’148 As with spinal opiates, there is an impression that breathing problems are more likely to occur in patients who are “narcotic naive.“141 Respiratory depression is reversible with intravenously administered naloxone, with minimal loss of analgesia.147 Tolerance is uncommon, and mild when it occ~rs.~~~~‘~~There is a small risk of infection that increases with time.153 NJWROAUGMENTATIVE

APPROACHES

GATE CONTROL THEORY OF PAIN Transmission of painful stimuli from peripheral nerve endings involves activation of small unmyelinated C fibers (slow fibers: 0.5 to 2 msec), which are present in great numbers and mediate vague, aching pain, and larger A delta fibers (fast transmission: 6 to 30 m/set), which are fewer in number, are more discrete, and transmit “pricking” pain. The gate control theory of Melzack and Wall provides a theoretical basis for the efficacy of stimulation-analgesia.‘54 The gate control theory attempts to explain pain perception on the basis of (1) a continuously changing equilibrium between converging input from small and large peripheral fibers; (2) modulation of impulses at the level of superficial laminae of the dorsal horn of the spinal cord (substantia gelatinosa); and (3) descending influences from higher CNS control centers. Peripheral activation of C fibers by low-threshold, non-noxious stimuli is postulated to inhibit painful sensations transmitted by A delta fibers through interaction at the level of the spinal cord. Barrages of peripheral stimuli are postulated to “close” a conceptual spinal “gate,” inhibiting the passage of concurrent nociceptive impulses and altering the quality of the pain message. The theory helps explain observations of pain relief associated with the application of massage, acupuncture, hydrotherapy, behavioral sensory altering techniques, and electrical stimulation. It is known that in postherpetic neuralgia, large nerve fibers are preferentially de172

Curr

Probl

Cancer,

May/June

1989

stroyed by the herpes virus, and it has been suggested that, as a result, gates are left “open” to continuous bombardment by small fibers. Work demonstrating an absence of reversal of transcutaneous electrical nerve stimulation (TENS)-induced analgesia after intravenous injections of naloxone and saline lends support for a mechanism independent of a simple placebo effect or activation of the endogenous opioid system.‘55 ELECTRICAL.

STIMULATION

Investigators have applied electrical stimulation at various levels of the nervous system to relieve pain through interaction with endogenous neuromodulatory mechanisms. Advantages of stimulationanalgesia include decreased dependence on narcotic drugs and avoiding the sequelae sometimes associated with neuroablative procedures. Electrical stimulation of the nervous system requires a pulse generator, an amplifier, and paired electrodes. Depending on the targeted site, the technique and complexity of instituting stimulation vary. Some systems are designed to enable patients to control the frequency and intensity of stimuli within preset limits, in response to varying analgesic requirements. TENS

Transcutaneous electrical nerve stimulation is the most commonly used stimulation modality. Rapid bursts of low-voltage electrical current are transmitted through a pair of electrodes applied to the skin overlying the painful region. At the proper settings, the stimulus is experienced as repeated painless paresthesias. Most systems use a rectangular waveform and permit regulation of current, frequency, and pulse width. Commercially available units are compact, simple to use, and relatively inexpensive. Numerous controlled studies of TENS have shown reduction of acute pain in over 60% of subjects and a long-term efficacy of 30% in reducing chronic pain. Other stimulation modalities have not been studied as thoroughly. Acute pain syndromes that respond well to TENS include rib fractures,156 postoperative incisional pain, *57-161labor, and dental pain.16’ For chronic pain, TENS is utilized in combination with other modalities (pharmacologic, behavioral, anesthetic), since the characteristic response is reduction rather than elimination of pain. Examples of chronic pain syndromes in which TENS has been used successfully include peripheral nerve injury, neuralgia, radiculopathy, and compression syndromes. Reports of results with TENS for pain related to malignant neoplasm are less encouraging. Ventafridda et al.163 obtained good to Curr

Probl

Cancer,

May/June

1989

173

excellent results in 43% of patients with cancer pain but noted significant reduction in efficacy with time, a characteristic observation.163 In a follow-up study, 96% of 37 patients with cancer pain noted marked reductions in pain intensity during the first 10 days of treatment, but by the end of 30 days this number had declined to 27% *163 Localized pain is more likely to respond than is vague, generalized pain. *64,165 Pain that is of central, psychogenic, or visceral origin is unlikely to respond we11.164 The individual response to stimulation varies. Trials for several weeks are necessary in some patients before maximum pain relief is realized.‘6g Transcutaneous electrical nerve stimulation is extremely safe and warrants trial as an adjunctive modality in patients with cancer pain. The main limitation of peripheral stimulation is a tendency for loss of effect with time.‘63 DORSAL

COLUMN

STIMULATION

Dorsal column stimulation involves the placement of stimulating electrodes in proximity to the spinal cord, in either the epidural or subarachnoid space. Historically laminectomy was required to implant electrodes, exposing patients to the rigors of general anesthesia and major surgery. New systems have been devised that are implantable through a standard 14-gauge epidural needle. Percutaneous insertion is accomplished by an anesthesiologist or neurosurgeon and has the advantage of enabling a trial of stimulation before the unit’s receiver is implanted subcutaneously. The main indications for dorsal column stimulation have been chronic back and leg pain, particularly for the “failed back’ or postlaminectomy syndrome.16’j Patients with multiple sclerosis, diabetic neuropathy, peripheral vascular disease, and stump neuromata have also been treated with some success.‘67-16s Strict guidelines for patient selection are still not established. Authors suggest that patients who demonstrate excessive pain behavior are less likely to benefit, and psychological screening is recommended.17” I71 Effective dorsal stimulation, like peripheral stimulation, requires that stimulation generate paresthesias over the painful region.170 There has been considerable experience with dorsal column stimulation instituted for the relief of cancer pain. Although many instances of success have been reported, overall results have been poor. 170,172J173 Pooled results should not be interpreted too strictly, since criteria for technique, follow-up, and success vary considerably, but a review of 16 series reporting on spinal stimulation for 88 patients with cancer pain indicates that treatment was successful in only 48% of patients.170’172 Increasing utilization of percutaneous methods of implantation, which are associated with reduced mor174

Curr

Probl

Cancer,

May/June

1989

bidity, may, in the future, permit sufficient to determine reliable screening criteria.

trials in cancer patients

DEEP Bl?AlN STIMULATION Chronic stimulation of deep brain structures is a recently introduced neurosurgical technique for the control of generalized pain. Deep brain stimulation is expected to play a prominent role in the management of widespread intractable cancer pain and may be of limited use for selected cases of nonmalignant pain. Reports are promising, although an insufficient number of patients h+ve been treated to determine long-term efficacy. Interest has focused on two distinct areas of the brain that, when stimulated, seem to produce pain relief by different mechanisms. The effects vary according to the area that is stimulated, which is taken into account when patients are selected.

Periventricular

and Periaqueductal

Gray Stimulation

Stimulation of these regions, as well as the medial posterior thalamus, is associated with the release of endogenous opioidlike substances (endorphins) into the third ventricle.‘74,175 Endorphins are hypothesized to activate a descending pain-inhibitory system originating in the central brain stem and terminating in the substantia gelatinosa, where suppression of pain transmission is postulated to occur. Pain relief resembles that which accompanies exogenous opioid administration, in that analgesia is reversed by naloxone administration and tolerance can occur with chronic stimulation. Tolerance can sometimes be reversed with disulfiram, tryptophan, or tricyclic compounds.176 The main advantage of periventricular stimulation over ablative procedures is the prospect of pain relief without the risks associated with major surgery or denervation procedures. Analgesia tends to be widespread and bilateral, rather than limited to a distinct area subserved by a single nerve. These characteristics make periventricular stimulation an attractive alternative for the treatment of pain that crosses the midline or involves the head, neck, or upper extremity. Destructive procedures designed to relieve these syndromes have low success rates, are associated with high incidences of mortality and morbidity, and may be accompanied by functional loss. According to the Young and Brechner177 review of periventricular and periaqueductal gray (PVG-PAG) stimulation for benign and malignant pain, 65% to 70% of carefully selected patients experienced pain relief sufficient to allow discontinuation of narcotic use and resumption of normal activities.‘77 Curr

Pmbi

Cancer,

May/June

1989

176

Stimulation of the Posterior Internal Capsule/Sensory Thalamus Stimulation of the internal capsule and discrete thalamic centers is being investigated for relief of central pain disorders. In contrast to stimulation of the PVG-PAG, endorphins are not released when these regions are stimulated, nor does tolerance occur. Pain relief is strictly contralateral and is not reversed when naloxone is administered. Distinct thalamic sensory relay nuclei are targeted for facial vs. body pain. Studies indicate that long-term pain control can be expected in greater than 50% of patients.‘7S1so Painful conditions that have been successfully treated include thalamic syndrome, anesthesia dolorosa, postherpetic neuralgia, spinal cord injury, brachial plexus avulsion, and phantom limb pain.l’l Despite promising results, the mechanism for pain relief is poorly understood. Theories include restoration of disrupted inhibitory impulses, increased sensory input to partially deafferented areas of somatosensory cortex, and interference with spontaneous neuronal hyperactivity.1828 ls3 PITUITARY

ABLATION

Luft and Olivecrona first advocated surgical hypophysectomy to reduce tumor spread in 1953.‘“a The observation that some patients experienced postsurgical pain relief led Moriccals4 to suggest pituitary destruction by percutaneous injection of absolute alcohol as a primary treatment for oncogenic pain. Alcohol ablation is preferred to surgical removal because it is relatively simple, safe, and inexpensive and entails only an abbreviated hospitalization. The incidence and severity of complications have been further reduced by technical modifications that permit selective destruction of the anterior hypophysis.185S lB6 This is accomplished either by limiting the quantity of injectate or introducing a cryoprobe and freezing tissue. Pituitary ablation is performed at a limited number of centers, in part because it is technically demanding. Clinical experience is sufficient for pituitary ablation to be regarded as an accepted pain-relieving modality. Pituitary destruction is considered for the treatment of intractable bilateral pain due to widespread bony metastases when life expectancy is moderate. Previously, selection of patients had been restricted to individuals with hormone-sensitive tumors. More recent reports suggest that the procedure is effective for patients with other malignant neoplasms and that localized head and neck pain may be particularly responsive.lB5’ lB7 Pituitary destruction per se does not adequately explain this procedure’s efficacy. The incidence of pain relief does not correlate with the degree of pituitary destruction implied by postoperative hormonal insufficiency.lB6 In a limited number of postmortem examinations of patients who had experienced good pain relief, only min176

Curr

Probl

Cancer,

May&me

1689

imal destruction of pituitary tissue was noted.‘% On initial injection, observers have noted cephalic migration of contrast material,lM lending support to the theory that pain relief is related to alcoholic destruction, thrombosis, and infarction of regions in the hypothalamus and thalamus, and further, visualization of dye within the third ventricle suggests that pathways within the forebrain may also be involved.‘ss Conversely, the success of isolated pituitary cryodestruction argues for a process originating in the gland itself, as does the reported efficacy of trials of temporary electrical stimulation to the gland?” Despite unchanged levels of metenkephalin and @endorphin measured in lumbar cerebrospinal fluid after pituitary ablation,lgO antanalgesia observed in many patients after naloxone administration suggests that interaction with the endogenous opiate system plays a role in pain relief, regardless of the primary target.l’l The technical aspects of the procedure have been carried out variously by nemosurgical or anesthesiologv teams, or in collaboration, with apparently similar results.“’ Once general anesthesia has been induced, a Is-gauge needle and introducer are passed through one nostril toward the pituitary fossa by the transsphenoidal route. The needle tip is localized just within the bony margins of the anterior part of the pituitary fossa with biplane fluoroscopy and image intensifier guidance. A ZO-gauge needle is advanced a few millimeters into the gland’s substance, and positioning is verified fluoroscopically after the injection of a minute quantity (0.1 ml) of contrast material. If alcohol is used, usually a total of 0.8 to 1.0 ml is injected in O.l-ml increments over 10 to 15 minutes. Alternatively, a cryoprobe is introduced through a 15-gauge needle, and a series of localized lesions are made within the gland. Four cryogenic lesions are created on each side of the gland, and 4 to 7 days later, four lesions are made in the midline. Before instilling alcohol or freezing, anesthesia is lightened to facilitate detection of pupillary movement or dilation. Pain relief usually develops gradually over 24 to 48 hours but in some cases occurs immediately. Hospitalization for a minimum of 48 hours is required to evaluate efficacy and alterations in analgesic requirements and to monitor for complications. Routine steroid replacement is undertaken for all patients not already receiving steroids. Results vary among studies owing in part to differences in patient selection and technique. In the six series reviewed, complete or almost complete pain relief was achieved in 42% to 98% of patients, 185,186,1s2-1s5 some of whom were injected more than one time. Other patients experienced lesser degrees of relief. In at least one case, analgesia was so profound as to allow a patient previously immobile from pain to return to work for an extended period.“’ In some series, osseous disease associated with breast and prostate cancer responded most favorably.185 There is no clear explanation Cut-r

Probl

Cancer,

May/June

1989

177

for the disparity in results in patients with non-endocrine-dependent tumors, if it indeed exists. The recent series of Lahuerta et al.ls7 included equal proportions of patients with hormone-dependent vs. non-hormone-dependent disease, and both groups fared equally we11.1s7 There is strong evidence that intervention should not be restricted to patients with hormone-dependent disease.187’18s”s2 In the series reviewed, the median duration of pain relief ranged between 7 weeks and 4 months. In some cases the degree of pain relief increased with the passage of time. The procedure has been repeated successfully in cases of early and late failure. Some patients remained free of pain for up to 2 years, and a significant number of patients died painlessly. MilesIs reported six procedure-related deaths in an early series of 250 procedures, but in a recent, more representative study, there was no mortality attributable to the procedure.187 The most frequent complications in a large series were headache (1721, diabetes insipidus 07%), and nausea (9%6).‘85 Headache and nausea were self-limited, but two thirds of patients with diabetes insipidus required vigorous therapy with antidiuretic hormone. In a recent study of 18 patients treated exclusively with cryotherapy, diabetes insipidus was not observed.ls5 Less common complications include epistaxis, meningitis, leakage of cerebrospinal fluid, and myxedema. Hypophysectomy is not ordinarily associated with alterations in neurologic or psychological function. Visual disturbance is a serious complication, related to damage to the optic chiasm. The most common finding is blurred vision, but permanent partial blindness has been reported. Incorrect needle placement and rapid injection of alcohol have been implicated as causes; a decreased incidence has been observed with the incremental injection of small volumes. Pupillary changes during the procedure correlate with postoperative visual disturbances, but their absence does not guarantee preservation of vision. NEUROSURGICAL GENERAL

INTERVENTION

CONSIDERATIONS

Ready access to neurosurgical opinion and intervention is an essential element in the successful implementation of a multidisciplinary cancer pain management program. With the exception of wellscreened cases of cranial neuralgias and advanced sympathetic dystrophy, the utility of neurosurgical pain procedures is primarily restricted to patients with intractable symptoms related to cancer whose life expectancy is limited. Even in this select group of patients, the need for neurosurgical intervention has been reduced by the relative success of refinements in pharmacologic therapy and 178

Cur-r

Probl

Cancer,

May/June

1989’

the advent of multidisciplinary care. These and other advances have increased the available alternatives to neurosurgery but at the same time have more clearly defined its role. As with chemical neurolysis, neurosurgical intervention is primarily reserved for patients who have failed thorough attempts at pharmacologic and other means of conservative control either because of insufficient analgesia or intolerance of side effects. Many neurosurgical procedures are analogous to anesthetic procedures and may require preoperative diagnostic-prognostic nerve blocks with local anesthetics. For complex pain problems, neurosurgical and anesthetic interventions can be complementary and together may provide more complete relief than either option would if it were applied alone. Finally, some procedures are not strictly circumscribed as neurosurgical or anesthetic and may be executed in collaboration (institution of epidural opiate infusions, dorsal column stimulation) or by either specialist alone (trigeminal neurolysis, pituitary ablation). Neurosurgical pain procedures can be generally classified as ablative (destruction of fibers by surgical section or thermal energy), augmentative (application of electrical stimulation), or delivery-oriented, as in the placement and maintenance of apparatus for CNS administration of opiate therapy (see “Spinal Opiates” section). Over a ?‘5year period, a variety of neurosurgical procedures have been introduced with the intention of providing safe, effective, and reliable pain relief (Fig 5). Most have been abandoned because these goals

@ DORSAL ROOT ENTRY ZONE LESION

@ MIDLI NE MYELf

\ 1 Lateral

(DAEZ)

Spinothalamic

FIG 5. Cross section of spinal cord indicating schematic sites for neuroblative procedures for pain control. (From Foley KM, Sundaresan M: Management of cancer pain, in Denita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology. Philadelphia, JB Lippincott, Co, 1984. Used by permission.) CUIT

Pmbl

Cancer,

May/June

1989

179

have not been adequately satisfied. The spectrum of analgesic neurosurgical operations has been extensively reviewed by one of us 196,197 md by OtherS.‘oo, 198,199General considerations pertinent to the application of neuroablative surgery are discussed below; because of its extreme clinical relevance, percutaneous cordotomy is described in detail. NEUROABZATIVE Limitations

SURGERY

and Hazards

Transience.-It is well established that chemical and surgical neurodestructive procedures rarely provide permanent relief of symptoms .loo The factors that influence the duration and quality of pain reiief are not well understood, but transience is consistent with current understanding of mechanisms of pain transm.ission. Within the context of a system that involves the activation of a multiplicity of diffuse pathways and gating mechanisms that allow patterns of activity to develop in widespread cortical and subcortical structures, it is not surprising that the simple interruption of a discrete neural tract does not reliably provide permanent relief.200 Deeply established patterns of pain behavior and central loci of pain help account for some treatment failures. Nevertheless, when life expectancy is limited by malignant neoplasm, transience is not a deterrent to well-planned surgery. Dysesthesia.-The frequency, severity, and persistence of central dysesthetic pain after ablative surgery varies considerably according to a number of factors.“’ Once established, most true central pain is relatively resistant to treatment efforts (see “Types of Pain” section), but because a latency period (analgesic window) is characteristic,100 this consequence is of limited concern in the preterminal patient. Stereotaxic destructive procedures (medial thalamotomy) and stimulation of the internal capsule have been advocated as last-resort remedies for deafferentation pain that is unresponsive to pharmacologic and behavioral measures .lgs,‘01 Damage to Nontargeted Structures.-Incidental damage to nontargeted structures can result in neurologic deficit and other injuries. The incidence of undesired neurologic deficit varies widely, even between series comparing identical procedures. Differences in technique, patient selection, criteria for deficit, and length of follow-up help to explain discrepancies. Modifications of surgical technique based on technologic advances (miniaturization, microscopy, stereotaxis, radiofkquency lesioning) have reduced, though not eliminated, undesirable sequelae for certain procedures.86~202 For some procedures, specific deficits are almost inevitable, whereas in others 180

Curr

Probl

Cancer,

May/June

1989

complications are uncommon. When a neuroablative procedure is being considered, the likelihood of specific dysfunction and the consequences to the individual patient must be balanced against the degree of benefit that is likely to be derived. The patient and his family need to be involved in decision making. Procedures followed by a high incidence of sphincteric disturbance are usually avoided unless a colostomy and/or urinary catheter are already required. The additional burden of care contributes to preexisting dysfunction, depression, and demands on the family unit. Alternatively, if risks are low, or the informed terminal patient is willing to accept the tradeoff for relief from intractable pain and drug dependence, then these procedures should be considered. Some patients regard the prospect of surgery or an additional procedure with dread, whereas others with intolerable levels of pain and dysfunction accept the risk of additional neurologic dysfunction with equanimity. Anesthetic requirements are exacting. Risks are unique and specific to the given procedure, and patients are subject to anesthetic considerations related to advanced neoplasm, previous surgery, and antitumor therapy. Percutaneous Cordotomy Percutaneous cordotomy remains the most frequently utilized neurosurgical procedure for the relief of cancer pain, particularly for unilateral pain confined to the trunk or lower limb. The prototype procedure, open cordotomy (Spiller and Martin, 19121, involves cervical or thoracic laminectomy and near-complete section of the anterolateral quadrant of the spinal cord, usually using general anesthesia. Percutaneous cordotomy (Mullan, 1963) has largely supplanted the open surgical approach?03’204 extending the relevance of spinothalamic tractotomy to patients too ill to undergo open surgical section safely. The percutaneous approach is commonly employed even when predicted life expectancy is limited to weeks or days. It is a simple, safe, and effective operation accompanied by minimal surgical and psychological trauma. Percutaneous cordotomy produces a stereotaxically guided lesion in the lateral spinothalamic tract within the cord’s anterolateral quadrant, most commonly at the Cl-2 level. The targeted fibers transmit pain and temperature sensation originating distally from the opposite side of the body. Electrical current and, more recently, thermocoagulation have superceded radiation as the means for generation of lesions. Impedance measurement is now commonly used to verify topography, and some investigators have added a thermocouple to regulate lesioning.205 Percutaneous cordotomy is ideally performed in a radiologic suite equipped with C-arm controlled fluoroscopy and an image intensifier. Local infiltration anesthesia is utilized and little or no sedation Cum Prabl

Cancer,

May/June

1989

181

is administered, to preserve the quality of patient cooperation and response. Generally, a ZO-gauge spinal needle is introduced laterally near the mastoid process between the C-l and C-2 vertebrae. Once the location of the tip of the probe has been electrically and radiologically verified, current is applied incrementally. The extent of contralateral analgesia and the absence of ipsilateral weakness are ascertained by testing the patient between coagulations. Indications.-Cordotomy is ideally suited for the treatment of intractable unilateral lower-extremity pain. Analgesia with preservation of proprioception, tactile sensation, and motor strength ordinarily results in a minimum of dysfunction. An identical degree of differential blockade cannot be simulated by the application of local anesthetics, so accurate preoperative prognostic neural blockade is impossible. When more extensive lesioning is employed to produce higher levels of analgesia, the incidence of both inadequate pain relief and complications increases. Perineal and abdominal wall pain, as well as pain originating from within the abdominal cavity, can usually be treated effectively without problemszo6 As reviewed by Ventafridda et al.,‘07 results are only fair for unilateral low brachial plexus and chest wall pain. The role of cordotomy is further limited in upperextremity pain because of a high failure rate and the increased incidence of respiratory and other neurologic complications. Pain involving both lower extremities or pelvic or back pain that crosses the midline is amenable to bilateral cordotomy, but bilateral procedures are now performed less frequently because of recognition of increased risks of paraparesis and bladder and respiratory dysfunction. Results and Complications.-Results are difficult to compare because of variations among series in patient selection, surgical technique, and follow-up. Immediate relief of malignant pain is usually cited as ranging between 60% and 80% .203~208 When failures are subjected to repeated cordotomy, overall success is rated at 86% to 96% ~205,208In one typical series of cancer patients treated with percutaneous cordotomy, 75% were pain free until their death. An additional 8% had significant pain relief, 8% had partial, transient relief, and 9% were considered failures.2o8 Homer’s syndrome restricted to the side operated on is invariably present in the case of successful cordotomy, as is thermoanesthesia corresponding to the zone of analgesia.2o8 Headache occurs in a large percentage of patients and responds to minor analgesics. Careful patient screening is essential since pain relief is transient, rarely persisting in excess of 1 to 2 years.208~20sOther limiting factors include the development of persistent dysesthesias in a variable per182

Cur-r Probl

Cancer,

May/June

1989

centage of patients’l’ and risks of neurologic dysfunction and ventilatory failure. The severity of dysesthesia is variable, and its incidence depends on how information is elicited. In a representative study, 40% of patients admitted to mild dysesthesia when questioned directly, 5% complained spontaneously, and 1% had burning dysesthesias that exceeded the severity of their original pain.210 The majority of patients with moderate life spans who undergo cordotomy will develop dysesthesias within 6 months. Discrete unilateral interruption of the spinothalamic tract should not result in loss of tactile or proprioceptive sensation, although unintentional destruction of neighboring structures can result in functional deficits. Efferent motor impulses are carried in the cerebrospinal tract of the posterolateral quadrant, which is separated from sensory tracts by the dentate ligament. The incidence of early ipsilateral limb weakness approaches 60% to 70% in most studies, but all but a few percent resolve within 1 month?08 Damage to the spinocerebellar tracts, which also occupy the anterolateral quadrant, produces ataxia in an estimated 0.5% of cases,2” although this complication is easily overlooked in the presence of motor weakness. Nerve fibers governing bladder function lie close to the lateral horn and may be interrupted, but bladder function is usually not impaired (1.5%) unless cordotomy is performed bilaterally, in which case some impairment is the rule.205~208 Ascending and descending respiratory reticular fibers lie near the anterior horn cells at the Cl-2 level. Derangement of respiratory fibers or interference with the phrenic outflow (C3-51 can contribute to respiratory difficulty, especially with bilateral procedures. Since sensory fibers are arranged in a laminar fashion with cervical bundles layered most anteriorly, extensive lesioning for more cephalad analgesia is more likely to be accompanied by respiratory insufficiency. Early mortality after cordotomy is commonly reported as ranging between 3% and 8% depending on patient selection, site of surgery, technique, and bilaterality.203’208 Several recent series have reported mortalities of less than 1% .'05,211 Almost without exception, deaths referable to the procedure itself are respiratory in nature. The association between respiratory death and bilateral cordotomy has been recognized since the 1930s.212 In 1962, Severinghaus and Mitchel1213 described the characteristic syndrome, labeling it Ondine’s curse. Ondine’s curse is a sleep apnea syndrome characterized by adequate ventilation during the awake state and respiratory failure during sleep, although apnea has also been observed during waking hourszo4 Other features sometimes present, include hypotension, bradycardia, generalized vasomotor instability, and SL4DH.202 The Curr

Probl

Cancer,

May/June

1989

183

syndrome usually appears within 24 hours of surgery, but its onset may be delayed by up to 1 week. When postoperative respiratory dysfunction occurs, overall mortality is 50% .‘14 Survivors’ symptoms have resolved within 3 to 32 days of their onset.“’ If there is no spontaneous resolution, death has so far been inevitable. The longest reported survivor died after 14 months despite the implantation of bilateral phrenic nerve pacemakers.‘15 The majority of respiratory deaths occur in individuals with preoperative pulmonary impairment, especially when cordotomy is performed for pain associated with bronchogenic carcinoma. Postoperative respiratory dysfunction following unilateral high cordotomy is uncommon in patients with normal preoperative pulmonary function. Bilateral cordotomy above C-4 carries a high risk of central apnea and death, so when surgery is considered for bilateral pain the safest course is to combine standard percutaneous cordotomy with contralateral open thoracic tractotomy, subarachnoid neurolysis, or low percutaneous lesioning.208 When bilateral lesioning is necessary, it should be planned in two stages separated by an interval of at least 1 week. BEHAVIORAL

ZNTERVENTZON

Even independent of physical symptoms, living with incurable cancer is associated with profound behavioral changes based just on the knowledge that terminal disease is present. Patients are forced to cope suddenly with impending loss of their image of good health, goals, dreams, loved ones, income, life-style, and even life itself. Reactions to these threats to survival take a myriad of forms, including sadness, regret, anger, and guilt. Fear and uncertainty produce anxiety, and depression is common. These reactions are usually exacerbated by the onset of symptoms, including pain, debility, dyspnea, nausea, constipation, and confusion. As symptoms introduce true alterations in life-style and behavior, a pervasive sense of helplessness and hopelessness frequently supervenes. Behavioral intervention formalizes and enhances the innate abilities of human beings to cope with adversity. In general, behavioral intervention involves instruction in specific techniques and skills that, with practice, the patient can utilize independently or with supervision to enhance the effectiveness of intercurrent methods of pain relief or, in some cases, defeat pain altogether. As early as 1914, Behan?16 in his textbook Pain, noted that ‘violent anger or great joy preempts the sensorium to such an extent that sense perception is dulled and may become absolutely negated.” Many of the modern behavioral techniques of pain and symptom control utilize similar strategies to distract the subject from pain. Techniques in common use include relaxation training, which may focus on the quality of breathing, guided imagery, self-hypnosis, cognitive therapy, operant 184

Curr

Probl

Cancer,

May/June

1989

therapy, and auditory distraction via headphones. Treatment can be facilitated by the use of prerecorded tapes, biofeedback to objectify end points, and a focus on muscle tension to reduce spasm. Selection of a technique is governed by a number of factors, including patient preference. The reader is referred to recent reviews for details on treatment.217’ ‘I8 Patient acceptance improves when the distinction between behavioral intervention and psychotherapy is clear, and when the decision to institute behavioral methods is understood not to reflect belief that pain is psychogenic. Treatment is more difficult when confusion is present or when concentration is impaired in association with the primary disease or intercurrent pharmacologic management. In contrast to “existential” anxiety, which has been described, anxiety may be primarily situational, in reaction to specific economic or social dilemmas. Identification and resolution of issues of concern to the patient often require great sensitivity. Members of the hospice care team are uniquely suited to intervene in these circumstances. CONCLUSION In general, chronic pain poses one of the greatest challenges in contemporary medical practice. In no group is the psychosocial impact of pain greater than in patients suffering with unrelieved oncologic pain. Health care providers who are interested in symptom management for this patient population are in a unique, exciting, and challenging position. Technological and conceptual advances within the various subspecialties have increased treatment options, and diverse tools, at various stages of development, are available to ensure patient comfort and dignity. These advances are evolving against a backdrop of heightened global awareness of the cancer pain problem and commitment at multiple levels to its resolution. Effective management demands commitment, compassion, and imagination, as well as currency of knowledge on the part of providers of primary and specialty care. The well-informed physician is in a critical position to provide essential educational guidance to the patient-family unit, the community at large, and other health care workers. Anesthesiologists are well suited to play an active part in the battle against cancer pain in virtue of their knowledge of clinical phannacology, pharmacokinetics, and pharmacodynamics, their skill in nerve-blocking techniques, and their involvement in the multidisciplinary delivery of pain treatment. The committed anesthesiologist can anticipate stimulating dialogue with colleagues from other specialties, an enhanced image in the community and hospital, and the most sincere gratitude from patients and their families. Curr

Probl

Cancer,

May/June

1989

186

REFERENCES 1. Racle JP, Abdellatif B, Poy JY, et al: Prolongation of isobaric bupivacaine spinal anesthesia with epinephrine and clonidine for hip surgery in the elderly. Anesth Analg 1987; 66M2-446. 2. Miralles FS, Lopez-Soriano F, Perez D, et al: Subarachnoid administration of calcitonin produces postoperative analgesia. Anesthesiology 1986; 65AlSS. 3. Carli P, Escoffey C, Chrubasik J, et al: Spread of analgesia and ventilatory response to CO, following epidural somatostatin. Anesthesiology 1986; 65~4216. 4. Payne R, Max M, Inturrisi C, et al teds): Principles ofAnalgesic Use in the Treatment ofAcute Pain and Chronic Cancer Pain. Washington, DC, American Pain Society, 1986. 5. Joranson DE, Dahl JL, Engber D: Wisconsin initiative for improving cancer pain management: Progress report. J Pain Symptom Management 1987; 2:111-114. 6. Kanner RM, Foley KM: Patterns of narcotic drug use in a cancer pain clinic. Ann NYAcad Sci 1981; 362:161-172. 7. Porter J, Jick H: Addiction rare in patients treated with narcotics. N Engl J Med 1980; 302:123. 8. Memkey H: Classification of chronic pain. Pain 1986; S3:226. 9. Butler S: Personal communication, May 1988. 10. Saunders C: The Management of Terminal Illness. London, London Hospital Medical Publications, 1967. 11. Foley KM: The management of pain of malignant origin, in Tyler HR, Dawson DM teds): Current Neurology. Boston, Houghton Mifflin, 1979, ~012, pp 279-302. 12. Daut RL, Cleeland CS: The prevalence and severity of pain in cancer. Cancer 1982;50:1913-1918. 13. Bonica JJ: Treatment of cancer pain: Current status and future needs. Adv Pain Res Ther 1985; 9:589-616. 14. World Health Organization: Cancer Pain Relief Geneva, World Health Organization, 1986. 15. Levin DN, Cleeland CS, Dar R: Public attitudes towards cancer pain. Cancer 1985;562337-2339. 16. Charap AD: The knowledge, attitudes, and experience of medical personnel treating pain in the terminally ill. Mt Sinai J Med 1978; 45561-580. 17. Bonica JJ: Management of Pain. Philadelphia, Lea &I, Febiger, 1954, pp 67% 701. 18. Black RG: Evaluation of the patient with the complaint of pain, in Raj PP ted): Practical Management ofPain. Chicago, Year Book Medical Publishers, 1986, pp 123-133. 19. National Institutes of Health Consensus Panel: Integrated approach to the management of pain. J Pain Symptom Management 1987; 2:35-44. 20. Miser AW: Assessment and treatment of pain in children with cancer. AnesthProg 1987;34:116-118. 21. Foley KM: Treatment of cancer pain. N Engl J Med 1985; 313:84-95. 22. Foley KM: Cancer pain syndromes. J Pain Symptom Management 1987; 2:T3-T7. 23. Foley KM: Pain syndromes in patients with cancer. Adv Pain Res Ther 1979; 2:.59-75. 24. Galasko CSB: Skeletal Me&stases. London, Butterworths, 1986, pp 99-124. 166

Curr

Probl Cancer, May/June

1989

25. Edeiken J, Karasick D: Imaging in cancer. CA 37239-245. 26. Gilbert RW, Kim JH, Posner JB: EpiduraI spinal cord compression from metastatic tumor: Diagnosis and treatment. Ann Nemo1 1978; 3:40-51. 27. Cousins MJ: Introduction to acute and chronic pain: Implications for neural blockade, in Cousins MJ, Bridenbaugh PO teds): Neural Blockade, ed 2. Philadelphia, JB Lippincott Co, 1988, pp 739-752. 28. Takeda F: Preliminary report from Japan on results of field testing of WHO draft interim guidelines for relief of cancer pain. Pain Clin 1986; 1:83-89. 29. Bonica JJ: Status of pain research and therapy. Semin Anesth 1986; 5:82-99. 30. Ferrer-Brechner T, Ganz P: Combination therapy with ibuprofen and methadone for chronic cancer pain. Am J Med 1984; 77178-83. 31. Beaver WI: Maximizing the benefits of weaker analgesics, in IASP RejFesher Courses on Pain Management. Hamburg, West Germany, International Association for the Study of Pain, 1987, pp l-25. 32. Baines M, Kirkham SR: Carcinoma involving bone and soft tissue, in WaII PD, Melzack R teds): Textbook ofPain. New York, Churchill Livingstone, Inc, 1984, pp 453-459. 33. OsteoIytic metastases, editorial. Lancer 1976; 2:1063-1064. 34. GaIasko CSB: Mechanisms of bone destruction in the development of skeletaI metastases. Nature 1976; 263:507-510. 35. Ferrer-Brechner T: Rational management of cancer pain, in Raj PP ted): Practical Management ofPain. Chicago, Year Book Medical Publishers, 1986, pp 312-328. 36. Kaiko RF, WaBenstein SL, Rogers AG, et ak Sources of variation in analgesic responses in cancer patients with chronic pain receiving morphine. Pain 1983; 15:191-200. 37. Cleeland CS, Tearnan BH: Behavioral control of cancer pain, in Holzman AD, Turk DC teds): Pain Management. New York, Pergamon Press, 1986, pp 193-212. 38. Twycross RG: Medical treatment of chronic cancer pain. Bull Cancer 1980; 67209-216. 39. PaaIzow LK: Pharmacokinetic aspects of optimal pain treatment. Acta Anaesthesiol Stand Suppl 1982; 74137-43. 40. Hanks GW: Antiemetics for terminal cancer patients. Lancet 1982; 1:1410. 41. Fosburg MF, Crone RK: Nitrous oxide analgesia for refractory pain in the terminaIIy ill. JAMA 1983; 2.50511-514. 42. Kaiko RF, Foley KM, Grabinski PY, et ak Central nervous system excitatory effects of meperidine in cancer patients. Ann Neuroll983; 13:180-185. 43. Inturrisi CE: Newer methods of opioid drug delivery, in IASP Refresher Courses on Pain Management Book of Abstracts. Hamburg, West Germany, International Association for the Study of Pain, 1987, vol 1, pp 27-39. 44. Nahata MC, Miser AW, Miser JS, et al: Analgesic plasma concentrations of morphine in children with terminal malignancy receiving a continuous subcutaneous infusion of morphine sulfate to control severe pain. Pain 1984; 18:109-114. 45. Magni G, Arsie D, De Leo D: Antidepressants in the treatment of cancer pain: A survey in Italy. Pain 1987; 29:347-353. 46. Mount BM: Psychological and social aspects of cancer pain, in Wall PD, Melzack R teds): Textbook of Pain. New York, Churchill Livingstone, Inc, 1984, pp 460-471. 47. Bond MR, Pearson IB: Psychological aspects of pain in women with advanced cancer of the cervix. J Psychosom Res 1969; 13:13-19. 48. Woodford JM, Fielding JR: Pain and cancer, in Weisenberg M (ed): Pain, Curr

Probl

Cancer,

May/June 1989

187

49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

70.

188

Clinical and E,xperimental Perspectives. St Louis, CV Mosby Co, 1975, pp 332-336. Sternbach RA, Timmermans G: Personality changes associated with reduction of pain. Pain 1975; 3:177-181. Kissen DM: The influence of some environmental factors on personality inventory scores in psychosomatic research.JPsychosom Res 1964; 8:145-149. Oxman T: A psychosomatic approach to the diagnosis of chronic pain, in Raj PP ted): Practical Management ofPain. Chicago, Year Book Medical Publishers, 1986, pp 134-145. Spiegel D, Bloom JR: Pam in metastatic breast cancer. Cancer 1983; 52:341345. Wand NG, Bloom VL, Friedel RO: Effectiveness of tricyclic antidepressants in treatment of coexisting pain and depression. Pain 1979; 7:331. Spiegel D, Bloom JR: Group therapy and hypnosis reduce metastatic breast carcinoma pain. Psychosom Med 1983; 45:333-339. Covi L: Drugs and group psychotherapy in neurotic depression. Am J Psychiatry 1974; 131:191-198. Halpern LM: Psychotropics and ataractics and related drugs. Adv Pain Res Ther 1979; 2275-283. Scott WA: The relief of pain with an antidepressant in arthritis. Practitioner 1969; 202:802-807. Watson CPN, Evans RI, Reed K, et al: Amitriptyline versus placebo in postherpetic neuralgia. Neurology 1982; 32:671-673. Jenkins DG, Ebbut AF, Evans CD: Imipramine in treatment of low back pain. J Int Med Aes 1976; 4(suppl):28-40. Max MB, Culnane M, Schafer SC, et al: Amitriptyline relieves diabetic neuropathy pain in patients with normal or depressed mood. Neurology 1987; 37:589-596. Trazadone (Desyrel): A new nontricyclic antidepressant. Med Lett Drugs Ther 1982; 24~47-48. Taub A, CoBins WF Jr: Observation on the treatment of denervation dysesthesia with psychotropic drugs. Adv Neural 1974; 4:309-316. Sweet WH: Treatment of trigeminal neuralgia (tic douloureux). N Engl J Med 1986; 315:174. Hatangdi VS, Boas RA, Richards EG: Postherpetic neuralgia: Management with antiepileptic and tricyclic drugs. Adv Pain Res Ther 1976; 1383-587. Swerdlow M: The use of anticonvulsants in the management of cancer pain, in Erdmann W, Oyama T, Pernak MJ teds): Pain Clinic. Utrecht, Netherlands, VNU Science Press, 1985. Lasagna RG, DeKornfeldt TJ: Methotrimeprazine: A new phenothiazine derivative with analgesic properties. JAMA 1961; 178:887-890. McGee JL, Alexander MR: Phenothiazine analgesia: Fact or fantasy. Am J Hosp Pharm 1979; 36:633-650. Forrest WH Jr, Brown BM Jr, Brown CR: Dextroamphetamine with morphine for the treatment of postoperative pain. N Engl J Med 1977; 13:712715. Faithful1 NS, Reinhold HS, Van Den Berg AP, et al: The effectiveness and safety of whole body hyperthermia as a pain treatment in advanced malignancy, in Erdmann W, Oyama T, Pernak MJ teds): Pain Clinic. Utrecht, Netherlands, VNU Science Press, 1985. Moricca G, Cavaliere R, Lopez M, et al: Combined whole body hyperthermia and chemotherapy in the treatment of advanced cancer with diffuse pain. Adv Pain Res Ther 1979; 2195-210. Curr Probl Cancer, May/June

1989

71. Mauch PM, Drew MA: Treatment of metastatic cancer to bone, in DeVita CT ted): Cancer: Principles and Practice of Oncology, ed 2. Philadelphia, JB Lippincott Co, 1985, pp 2132-2141. 72. Twycross RG: Relief of pain in far-advanced cancer. Reg Anaesth 1980. 73. Mellette SJ: Management of malignant disease metastatic to the bone by hormonal alterations. Clin Orthop 1970; 73:73-88. 74. Dogliotti AM: Traitement des syndromes doulomux de la peripherie par l’alcolisation sub-arachnoidienne des racines posterieures a leur emergence de la moelle epiniere. Presse Med 1931; 39:1249-1253. 75. Maher RM: Relief of pain in incurable cancer. Lancet 1955; 1:18. 76. Gerbershagen HLJ: Blocks with local anesthetics in the treatment of cancer pain. Adv Pain Res Ther 1979; 2:311-323. 77. Pilon RN, Baker AR: Chronic pain control by means of an epidural catheter. Cancer 1976; 37:903-905. 78. Du Pen SL: After epidural narcotics: What next for cancer pain control. Anesth Analg 1987; 66846. 79. Cousins MJ, Bridenbaugh PO teds): Neural Blockade, ed 2. Philadelphia, JB Lippincott Co, 1988. 80. Stovner J, Endresen R: Intrathecal phenol for cancer pain. Acta Anaesth Stand 1972; 16:17-21. 81. Raj PP ted): Practical Management of Pain. Chicago, Year Book Medical Publishers, 1986, pp 857-858. 82. Smith MC: Histological findings following intrathecal injections of phenol solutions for relief of pain. Br J Anaesth 1963; 36:387406. 83. Stewart WA, Lourie H: An experimental evaluation of the effects of subarachnoid injections of phenol-pantopaque in cats: A histological study. J Neurosurg 1963; 20:64-72. 84. Doyle D: Nerve blocks in advanced cancer. Practitioner 1982; 226:539-544. 85. DeWald R, Bridwell KH, Prodromas C, et al: Reconstructive spinal surgery as palliation for metastatic malignancies of the spine. Spine 1985; 1021. 86. Siegfried J, Broggi G: Percutaneous thermocoagulation of the gasserian ganglion in the treatment of pain in advanced cancer. Adv Pain Res Ther 1979; 2:463-468. 87. Madrid JL, Bonica JJ: Cranial nerve blocks. Adv Pain Res Ther 1979; 2:347355. 88. Simon DL, Carton H, Rowlingson JC: Treatment of bladder pain with transsacral nerve block. Anesth Analg 1982; 61:46-48. 89. Robertson DH: Transsacral neurolytic nerve block. Br J Anaesth 1983; 55:873-875. SO. Bonica JJ: The management of pain of malignant disease with nerve blocks. Anesthesiology 1954; 15:280-301. 91. Swerdlow M: Intrathecal neurolysis. Anaesthesia 1978; 33:733-740. 92. Peyton WI, Semansky EJ, Baker AB: Subarachnoid injection of alcohol for relief of intractable pain with discussion of cord changes found at autopsy. Am J Cancer 1937; 30:709. 93. Ischia S, Luzzani A, Ischia A, et al: Subarachnoid neurolytic block U-T) and unilateral percutaneous cervical cordotomy in the treatment of pain secondary to pelvic malignant disease. Pain 1984; 20:139-149. 94. Takagi Y, Koyama T, Yamamoto Y: Subarachnoid neurolytic block with 15% phenol glycerine in the treatment of cancer pain. Pain 1987; S4:T33. 95. Katz J: Current role of neurolytic agents. Adv Neurot 1974; 4:471476. 96. Swerdlow M: Subarachnoid and extradural blocks. Adv Pain Res Ther 1979; 2:325-337. Curr

Probl

Cancer,

May/June 1989

189

97. Holland AJC, Youssef M: A complication of subarachnoid phenol blockade. Anaesthesia 1979; 34260-262. 98. Swerdlow M: Spinal and peripheral neurolysis for managing Pancoast syndrome. Adv Pain Res Ther 1982; 4:135-144. 99. Pagni CA: Nemosurgical treatment: Status of the problem. Adv Pain Res Ther 1982; 4:165-182. 100. White JC, Sweet WH: Pain and the Neurosurgeon: A 40-Year E,xperience. Springfield, Ill, Charles C Thomas, Publisher, 1969. 101. DeBacker IJ, Kienzle WK, Keasling HH: A study of stellate ganglion block for pain relief. Anesthesiology 1959; 20:618-623. 102. Warfield CA, Crews DA: Use of stellate ganglion blocks in the treatment of intractable limb pain in lung cancer. C/in J Pain 1987; 3:13-15. 103. Mullin V: Brachial plexus block with phenol for painful arm associated with Pancoast’s syndrome. Anesthesiology 1980; 53:431-433. 104. Hay RC: Subarachnoid alcohol block in the control of intractable pain. Anesth Analg 1962; 41:12-16. 105. Watson CPN, Evans RJ: Intractable pain with cancer of the rectum. Pain Clin 1986; 129-34. 106. Rohde .I, Hankemeier U: Neurolytic caudal blocks for the relief of perianal cancer pain. Pain 1987; S4:T32. 107. Ferrer-Brechner T: Epidural and intrathecal phenol neurolysis for cancer pain. Anesthesiol Rev 1981; 8:14-19. 108. Racz GB, Sabonghy M, Gintautas J: Intractable pain therapy using a new epidural catheter. JAMA 1982; 24B:579-581. 109. Jain S, Foley K, Thomas J, et al: Factors influencing efficacy of epidural neurolysis therapy for intractable cancer pain. Pain 1987; S4:T34. 110. Shibutani K, Kizelshteyn G, Allyne L, et al: Low volume intermittent lumbar epidural phenol injection for relief of cancer pain. Pain 1987; S4:T32. 111. Racz GB, Heavner J, Haynsworth R: Repeat epidural phenol injections in chronic pain and spasticity, in Lipton S, Miles J teds): Persistent Pain. New York, Grune & Stratton, 1985, vol 5, pp 157-179. 112. Korevaar WC, Kline MT, Donnelly CC: Thoracic epidural neurolysis using alcohol. Pain 1987; S4:T33. 113. Miller RD, Munger WL, Powell PE: Chronic pain and local anesthetic neural blockade, in Cousins MJ, Bridenbaugh PO teds): Neural Blockade, ed 1. Philadelphia, JB Lippincott Co, 1980, pp 616-636. 114. Payne R: Clinical neuropathic pain syndromes, with special reference to causalgia and reflex sympathetic dystrophy. Clin J Pain 1986; 2:59-73. 115. Lofstrom JB, Cousins MJ: Sympathetic neural blockade of upper and lower extremity, in Cousins MJ, Bridenbaugh PO: Neural Blockade, ed 2. Philadelphia, JB Lippincott Co, 1988, pp 461-500. 116. Carron H, Litwiller R: Stellate ganglion block. Anesth Analg 1975; 54:567-570. 117. Winnie AP: Plexus Anesthesia. Philadelphia, WB Saunders Co, 1983. 118. Hatangi VS, Boas RA: Lumbar sympathectomy: A single needle technique. Br J Anaesth 1985; 57285-289. 119. Umeda S, Toshiyuki A, Hatano Y, et al: Cadaver anatomic analysis of the best site for chemical lumbar sympathectomy. Anesth Analg 1987; 66:643646.

120. Goucke CR, Lovegrove radiopharmaceuticals 1987; 59:944-945.

190

FTA, Finch PM: Lumbar sympathectomy: Spread of following a single needle technique. Br J Anaesth

Curr

Bob1

Cancer,

May/June 1989

121. Moore DC: Regional Block, ed 4. Springfield, Ill, Charles C Thomas, Publisher, 1975. 122. Ischia S, Luzzani A, Ischia A, et al: A new approach to the neurolytic block of the coeliac plexus: The transaortic technique. Pain 1983; 16333. 123. Singler RC: An improved technique for alcohol neurolysis of the celiac plexus. Anesthesiology 1982; 56:137-141. 124. Flanigan DP, Kraft R: Continuing experience with palliative chemical splanchnicectomy. Arch Surg 1978; 113:509-511. 125. Brown DL, Bulley CK, Quiel EC: Neurolytic celiac plexus block for pancreatic cancer pain. Anesth Analg 1987; 66:869-873. 126. Thompson GE, Moore DC, Bridenbaugh PO, et al: Abdominal pain and alcohol celiac plexus nerve block. Anesth Analg 1977; 56:1-5. 127. Jones J, Gough D: Coeliac plexus block with alcohol for relief of upper abdominal pain due to cancer. Ann R Co11 Surg Engll977; 59:46-49. 128. Benzon HT: Convulsions secondary to intravascular phenol: A hazard of celiac plexus block. Anesth Analg 1979; 58:150-151. 129. Yoshihisa F, Takaori M: Pleural effusion after CT-guided alcohol celiac plexus block. Anesth Ana@ 1987; 66911-912. 130. Galizea EJ, Lahiri SK: Paraplegia following coeliac plexus block with phenol. Br J Anaesth 1974; 46:539-540. 131. Fine PG, Bubela C: Chylothorax following celiac plexus block. Anesthesiology 1985;63:454-456. 132. Moore DC, Bush WI-I, Burnett LL: Celiac plexus block: A roentgenographic, anatomic study of technique and spread of solution in patients and corpses. Anesth Analg 1981; 60:369-379. 133. Ward EM, Rorie DK, Nauss LA, et al: The celiac ganglia in man: Normal anatomic variations. Anesth Analg 1979; 58:461-465. 134. Bromage PR, Camporesi E, Chestnut D: Epidural narcotics for postoperative analgesia. Anesth Analg 1980; 59:473-480. 135. Max MB, Inturrisi CE, Kaiko RF, et al: Epidural and intrathecal opiates: Cerebrospinal fluid and plasma protiles in patients with chronic cancer pain. Clin Phamacol Ther 1985; 38:631-641. 136. Pasternak GW: Multiple morphine and enkephalin receptors and the relief of pain. JAMA 1988; 259:1362-1367. 137. Woerd AVD, Oyama T, Trouwborst A, et al: Intrathecal B-endorphic treatment in man with intractable cancer pain. Pain Clin 1985; 199-106. 138. Qyama T, Jin T, Murakawa T: Analgesic effect of continuous intrathecal beta-endorphin in cancer patients. Pain Clin 1985; 1:93-98. 139. Abrams S: Current concepts in the use of epidural narcotics in post-operative, cancer and labor pain. ASA Refresher Course Lectures. American Society of Anesthesiologists, 1987, p 512. 140. Coombs DW, Saunders RS, Lachance D, et al: Intrathecal morphine tolerance: Use of intrathecal clonidine, DADLE, and intraventricular morphine. Anesthesiology 1985; 62:358-363. 141. Lobato RD, Madrid JL, Fatela LV, et al: Intraventricular morphine for control of pain in terminal cancer patients. J Neurosurg 1983; 59:627-633. 142. Coombs DW, Saunders RL, Gaylor MS, et al: Relief of continuous chronic pain by intraspinal narcotics infusion via an implanted reservoir. JAlMA 1983; 2502336-2340. 143. Pope K: A service station for neivous disorders. NY Acad Sci Focus 1987; 2:1-7.

Cur-r Pmbl

Cancer,

May/June 1989

191

144. Zenz M, Piepenbrock S, Tryba M: Epidural opiates: Long-term experiences in cancer pain. Klin Wochenschr 1985; 63:225-229. 145. Shetter AG, Hedley MN, Wilkinson B: Administration of intraspinal morphine sulfate for the treatment of intractable cancer pain. Neurosurgery 1986;18:740-747. 146. Schoeffler PF, Haberer JP, Monteillard CM, et al: Morphine injections in the cisterna magna for intractable pain in cancer patients. Anesthesiology 1987; 67A246. 147. Lazorthes Y, Verdie JC, Bastide R, et al: Spinal versus intraventricular chronic opiate administration with implantable drug delivery devices for cancer pain. Appl Neurophysioll985; 48234-241. 148. Lenzi A, Galli G, Gandolfini M, et al: Intraventricular morphine in paraneoplastic painful syndrome of the cervicofaciai region: Experience in 38 cases. Neurosurgery 1985; 176-U. 149. Obbens FAMT, Stratton Hill C, Leavens ME, et al: Intraventricular morphine administration for control of chronic cancer pain. Pain 1987; 28:61-68. 150. Roquefeuil B, Benezech J, Blanchet P, et al: Intraventricular administration of morphine in patients with neoplastic intractable pain. Surg Neural 1984; 21:155-158. 151. Nurchi G: Use of intraventricular and subarachnoid morphine in intractable pain associated with cancer. Neurosurgery 1984; 15:801-803. 152. Richardson DE, Akil H: Pain reduction by electrical brain stimulation in man: II. Chronic self-administration in the periventricular grey matter. J Neurosurg 1977; 47:184-194. 153. Black P: Neurosurgical management of cancer pain. Semin Oncol 1985; 12:438-444. 154. Melzack R, Wall PD: Pain mechanisms: A new theory. Science 1965; 150:971-979. 155. Abram SE, Reynolds AC, Cusick JF: Failure of naloxone to reverse analgesia from transcutaneous electrical stimulation in patients with chronic pain. Anesth Analg 1981; 60:81-84. 156. Myers RA, Woolf CJ, Mitchell D: Management of acute traumatic pain by peripheral transcutaneous electrical stimulation. S Afr Med J 1977; 52:309312. 157. Ali JA, Yatfee CS, Serretti C: The effect of transcutaneous electrical nerve stimulation on postoperative pain and pulmonary function. Surgery 1981; 89:507-512. 158. Cooperman AM, Hall B, Mikalacki K, et al: Use of transcutaneous electrical stimulation in control of postoperative pain: Results of a prospective randomized study. Am J Surg 1977; 133:185-187. 159. Pike PM: Transcutaneous electrical stimulation: Its use in management of postoperative pain. Anaesthesia 1978; 33:165-171. 160. Solomon RA, Viernstein MC, Long DM: Reduction of postoperative pain and narcotic use by transcutaneous electrical nerve stimulation. Surgery 1980; 87:142-146. 161. Vander Ark GD, McGrath KA: Transcutaneous electrical stimulation in treatment of postoperative pain. Am J Surg 1975; 130:338-340. 162. Augustinsson LE, Bohlin P, Bundsen P, et al: Pain relief during delivery by transcutaneous nerve stimulation. Pain 1977; 4:59-65. 163. Ventafridda V, Sganzerla EP, Fochi C, et al: Transcutaneous nerve stimulation in cancer pain. Adv Pain Res Ther 1979; 2:509-515. 164. Nielzen S, Sjolund BH, Eriksson BE: Psychiatric factors influencing the 192

Curr Probl Cancer,

May/June 1989

165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184.

treatment of pain with peripheral conditioning stimulation. Pain 1982; 13:365-371. Wall PD, Sweet WH: Temporary abolition of pain in man. Science 1967; 155:108-109. De La Porte C, Siegfried J: Lumbosacral spinal fibrosis (spinal arachnoiditis): Its diagnosis and treatment by spinal cord stimulation. Spine 1983; 8:593-603. Feeney DM, Gold GN: Chronic dorsal stimulation: Effects on H reflex and symptoms in a patient with multiple sclerosis. Neurosurgery 1980; 6564. Richardson RR, Siqueira EB, Cerullo LJ: Spinal epidural stimulation for treatment of acute and chronic intractable pain: Initial and long-term results. Neurosurgery 1979; 5:344. Fiume D: Spinal cord stimulation in peripheral vascular pain. Appr Neurophysioll983; 46290-294. Loeser JD: Dorsal column and peripheral nerve stimulation for relief of cancer pain. Adv Pain Res Ther 1979; 2:499-507. Long DM, Erickson D, Campbell J, et al: Electrical stimulation of the spinal cord and peripheral nerves for pain control: A 10 year experience. Appl Neurophysiol1981;44207-217. Meglio M, Cioni B: Personal experience with spinal cord stimulation in chronic pain management. Appl Neurophysioll982; 45:195-200. Shealy CN, Mortimer JT, Reswick JB: Electrical inhibition of pain by stimulation of the dorsal columns: A preliminary report. Anesth Analg 1967; 46:489-491. Akil H, Richardson DE, Hughes J, et al: Elevations in levels of enkephalinlike materials in the ventricular CSF of pain patients upon analgetic focal stimulation. Science 1979; 201:463-465. Richardson DE: Long-term follow-up of deep brain stimulation for relief of chronic pain in the human, in Brock M ted): Modern Neurosurgery. New York, Springer-Verlag New York, 1982, pp 449-453. Dieckmann G, Witzmann A: Initial and long-term results of deep brain stimulation for chronic intractable pain. Appl Neurophysiol 1982; 45:167172. Young RF, Brechner T: Electrical stimulation of the brain for relief of intractable pain due to cancer. Cancer 1986; 57:1266-1272. Hosobuchi Y: Subcortical electrical stimulation control of intractable pain: Report of 122 cases. 3 Neurosurg 1986; 64:543-553. Turnbull JM, Shulman R, Woodhurst WB: Thalamic stimulation for neuropathic pain. J Neurosurg 1980; 52486-493. Young RF, Kroening R, Fulton W, et aI: Electrical stimulation of the brain in the treatment of chronic pain: Experience over 5 years. J Neurosurg 1985; 62:389-396. Young RF: Surgical role in relief of pain. Semin Anesth 1985; 4:323-331. Adams JE: Electrical stimulation of the internal capsule for the control of central pain, in Morley TP ted): Current Controversies in Neurosurgery. Philadelphia, WB Saunders Co, 1976, pp 510-514. Andy OJ: Thalamic stimulation for chronic pain. Appl Neurophysiol 1983; 46:116-123. Morrica G: Pituitary neuroadenolysis in the treatment of intractable pain from cancer, in Lipton S ted): Persistent Pain: Modern Methods of Treatment. New York, Academic Press, 1977, vol 1, pp 149-173.

Cur-r Prwbl

Cancer,

May/June 1989

193

184a. Luft R, Olirecrona H: Experiences with hypophysectomy. J Neurosurg 1953; 10:301-316. 185. Duthie AM, Ingham V, Dell AE, et al: Pituitary cryoablation: Results of treatment using a transphenoidal cryoprobe. Anaesthesia 1983; 38d48-451. 186. Lloyd JW, Rawlinson WAL, Evans PJD: Selective hypophysectomy for metastatic pain: A review of ethyl alcohol ablation of the anterior pituitary in a regional pain relief unit. Br J Anaesth 1981; 53:1129-1132. 187. Lahuerta J, Lipton S, Miles J, et al: Update on percutaneous cervical cordotomy and pituitary alcohol neuroandenolysis: An audit of our recent results and complications, in Lipton S, Miles J (edsl: Persistent Pain. New York, Grune & Stratton, 1985, vol 5, pp 197-223. 188. Madrid JL: Chemical hypophysectomy. Adv Pain Res Ther 1979; 2:381-392. 189. Corssen G, Yanagida H, Trouwborst A, et al: Relief of cancer pain: Alcoholinduced neuroadenolysis vs. electrical stimulation of the pituitary gland. Pain Clin 1985; 1:136-141. 190. Miles J: Chemical hypophysectomy. Adv Pain Res Ther 1979; 2:373-380. 191. Richardson DE: Role of neurosurgery in pain involving the chest and brachial plexus. Adv Pain Res Ther 1979; 2:577-585. 192. Miles J: Pituitary destruction, in Wall PD, Melzack R teds): TeKtbook @Pain. New York, Churchill Livingstone, Inc, 1984, pp 656-665. 193. Corssen G, Holcomb MC, Moustapha I, et al: Alcohol-induced adenolysis of the pituitary gland: A new approach to control intractable cancer pain. Anesth Analg 1977; 56:414. 194. Gye RS, Stanworth PA, Stewart JA, et al: Cryohypophysectomy for bone pain of metastatic breast cancer. Pain 1979; 6201-206. 195. Lipton S: Percutaneous cervical cordotomy and the injection of the pituitary with alcohol. Anaesthesia 1978; 33953. 196. Patt R: Neurosurgical interventions for chronic pain problems. Anesthesiol Clin North Am 1987; 5609-638. 197. Patt R: Pain therapy, in Frost EAM: Clinical Anesthesia in Neurosurgery, ed 2. Boston, Butterworths, in press. 198. Long DM: Surgical therapy of chronic pain. Neurosurgery 1980; 6:317-326. 199. Bullitt E, Friedman A, Sekhar LN: Neurosurgical procedures, in Raj PP ted): Practical Management of Pain. Chicago, Year Book Medical Publishers, 1986, pp 728-760. 200. Pagni CA: General comments on ablative nemosurgical procedures. Adv Pain Res Ther 1979; 2:405-424. 201. Tasker RR: Stereotactic surgery, in Wall PD, Melzack R teds): Textbook of Pain. New York, Churchill Livingstone, Inc, 1984, pp 639-655. 202. Krieger A, Rosomoff H: Sleep induced apnea: I. A respiratory and autonomic dysfunction syndrome following bilateral percutaneous cervical cordotomy. J Neurosurg 1969; 31:620-627. 203. Lahuerta J, Lipton S, Wells JCD: Percutaneous cervical cordotomy: Results and complications in a recent series of 100 patients. Ann A Co11 Surg Engl 1985; 67:41-44. 204. Tasker RR: Merits of percutaneous cordotomy over the open operation, in Morley TF’ ted): Current Controversies in Neurosurgery. Philadelphia, WB Saunders Co, 1976, pp 496501. 205. Levin AB: Techniques and results of cordotomy in patients with pain of benign and mallgnant origin, in Rizzi R, Visentin M teds): Pain Therapy. Amsterdam, Elsevier Science Publishers, 1983. 206. Loeser JD: Role of neurosurgery in visceral and perineal pain. Adv Pain Res Ther 1979; 2:607-614. 194

Curr Probl Cancer, May/June 1986

207. Ventafridda V, De Conno F, Fochi C: Cetvical percutaneous cordotomy. Adv Pain Res Ther 1983; 4:185-198. 208. Lipton S: Percutaneous cordotomy, in Wall PD, Melzack R feds): Textbook of Pain. New York, Churchill Livingstone, Inc, 1984, pp 632-638. 209. Nathan Pw: Results of antero-lateral cordotomy for pain in cancer. J Neurol Neurosurg Psychiatry 1963; 26:353-362. 210. Mullan S: Percutaneous cordotomy. Adv Neural 1974; 4:677-682. 211. Ischia S, Luzzani A, MaBezzoli GF, et al: Percutaneous cervical cordotomy: Technical considerations and results in 400 treated cases, in Rizzi R, Visentin M teds): Pain Therapy. Amsterdam, Elsevier Science Publishers, 1983, pp 367-379. 212. Stookey B: Chordotomy of the second cervical segment for relief from pain due to recurrent carcinoma of the breast. Arch Neural Psychiatry 1931; 26443. 213. Severinghaus JW, Mitchell RA: Ondine’s curse: Failure of respiratory center automatic@ while awake. Clin Res 1962; 10:122. 214. Lipton S: Percutaneous cervical cordotomy. Adv Pain Res Ther 1979; 2:425437. 215. Polatty RC, Cooper KR: Respiratory failure after percutaneous cordotomy. South Med J 1986; 79:897-899. 216. Behan FiJ: Pain: Its Origin, Conduction, Perception and Diagnostic Sign$cance. New York, D. Appleton, 1921. 217. Sternbach RA: Specific therapeutic techniques in pain management. IASP Refresher Courses in Pain Management Book of Abstracts. Hamburg, West Germany, International Association For the Study of Pain, 1987; vol 1, pp 107-112. 218. Baines M: Practical care of terminally ill patients. Midlands Med 1986; 1:1721.

Curr Prebl Cancer, May/June 1989

196

Richard

B. Patt, M.D., is a board-certified anesthesiologist trained at the Albert Einstein College of Medicine and Memorial Sloan Kettering Hospital with fellowships in Pain Management and Neuroanesthesia. He is Assistant Professor of Anesthesiology and Psychiatry at the University of Rochester School of Medicine and Dentistry where he provides surgical anesthesia one day per week. The bulk of Dr. Patt’s energy is devoted to research and clinical care in Strong Memorial Hospital’s Multidisciplinary Pain Treatment Center in Rochester, New York. Dr. Patt has initiated and driven the Center’s Cancer Pain Service and Cancer Pain Initiative. Subhash Cornell and pain Hospital morial’s clinician, ment of

196

Jain, M.D., is an anesthesiologist on the faculty of Medical School and has been pro’oviding anesthesia management services at Memorial Sloan Kettering for more than 10 years. He is the Director of MeNerve Block Clinic, and is a respected researcher, and educator with special expertise in the managecancer pain.

Curr

Probl

Cancer,

May/June

1989