Opioid pharmacotherapy in the management of cancer pain. A survey ...

1283

Opioid Pharmacotherapy in the Management of Cancer Pain A Survey of Strategies Used by Pain Physicians for the Selection of Analgesic Drugs and Routes of Administration Nathan 1. Cherny, M.B.B.S., F.R.A.C.P., Victor Chang, M.D., Gerri Frager, M.D., J a n e M . Ingharn,M.B.B.S., F.R.A.C.P., Paul]. Tiseo, Ph.D., Beth Popp, M.D., Russell K. Portenoy, M.D., and Kafhleen M . Foley, M.D.

Background. This survey documents the strategies used by pain control physicians in the selection of opioid drugs and routes of administration in the management of inpatients referred to a cancer pain service. Methods. The following approaches were prospectively evaluated during the treatment of 100 consecutive inpatients: 1) the influence of the evaluation of the goals of care on decision making, 2 ) selection of opioid drugs, 3) indications for changing opioid drugs and the frequency with which this strategy is used, and 4)selection of route of administration. Results. Eighty of the 100 patients underwent a total of 182 changes in drug, route, or both drug and route before discharge or death. The major reasons for change were to improve the convenience of treatment regimen in the setting of adequate pain relief (31.4%), diminish side effects in the setting of controlled pain (25.0%), reduce the invasiveness of therapy in the setting of controlled pain (19.3%), and simultaneously improve pain control and reduce opioid toxicity (17.7%). When opioid toxicity was the reason for change, physicians changed the opioid drug in 71% of cases and the route in 29%. When convenience or invasiveness were targeted, the physicians changed the route in 61% of cases and the opioid in 39%. Forty-four patients required one or more change in the opioid, and 20 required 2 or more changes (range, 2-6 changes). At the time of discharge (n = 82), morphine was more commonly selected than hydromorphone or fentanyl(39% vs. 23% vs. 17%) and the routes of administration were oral (57%), transdermal (l8%), intravenous (l8%), subcutaneous IS%), and intraspinal (4%). TheraFrom the Pain Service, Department of Neurology, Memorial Sloan-Kettering Cancer Center, New York, New York. Supported in part by National Cancer Institute grant CA-32897. Address for reprints: Kathleen M. Foley, M.D., Pain Service, Department of Neurology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. Received May 10,1995; accepted May 24, 1995.

peutic changes were associated with improvement in physician-recorded pain intensity and a lower prevalence of cognitive impairment, hallucinations, nausea and vomiting, and myoclonus among patients who were discharged from the hospital. Conclusions. These data illustrate the application of strategies for selections of opioid drugs and their route of administration that are recommended in current guidelines for the management of cancer pain. Cancer 1995;76: 1288-93. Key words: cancer pain, opioid therapy, decision making, palliative treatment.

In recent years, guidelines for the management of cancer pain have been issued by various experts'-3 and organization~.~-~ These guidclines have emphasized opioid pharmacotherapy as the mainstay of therapy. Although early guidelines emphasized the use of morphine administered more recent guidelines have emphasized the individualization of therapy, which is accomplished by the selection of the most appropriate drug and route of administration for the specific clinical c i r c u m s t a n c ~ . ' According ~ ~ ~ ~ ~ ~ to these guidelines, the aims of therapy are to achieve adequate pain relief safely within an acceptable time frame, to minimize the side effects of treatment, and to provide ongoing analgesic therapy by the most convenient and least noxious means available. Despite the dissemination of these guidelines, recent surveys of physicians8 and their patients' have indicated that clinicians continue to have difficulty in selecting analgesics and routes of administration during the long term management of cancer pain. Evidence suggests that guidelines alone are unlikely to change clinical practice.'"

1284

CANCER October 1,1995, Volume 76,No. 7

The effectiveness of guidelines may be enhanced through a detailed description of the decision-making process implemented during the treatment of illustrative case material. This description can demonstrate the rationale and outcomes of specific decisions, thereby clarifying the practical considerations that must be addressed as guidelines are applied in practice. Because therapeutic strategies may evolve during the patient's disease course, the most informative description should incorporate a longitudinal perspective. To review the strategies that are used in selecting an opioid and route of drug administration during the management of chronic cancer pain, we prospectively evaluated the practices of Pain Service physicians at Memorial Sloan-Kettering Cancer Center. Data were collected by the treating physician, who recorded the reason(s) for referral and change(s) in the analgesic regimen. Methods

The Pain Service at Memorial Sloan-Kettering Cancer Center is an interdisciplinary clinical and research service for the study of cancer pain that was established in the department of neurology 18 years ago. Participating clinicians include physicians (neurologists, oncologists, and palliative medicine physicians), nurses, a clinical pharmacist, and a social worker. The Pain Service provides care for inpatients and outpatients referred for problems with symptom control, predominantly pain. Inpatients referred to the Pain Service are initially evaluated by physician Fellows. The aims of the evaluation are to identify the pain syndrome and predominating pathophysiologic mechanism; to assess the severity of the pain, the patient's response to previous trials of therapy, and the pain's effect on physical, emotional, and psychologic function; the extent of underlying disease; and the prevailing goals of patient care. Most patients undergo diagnostic tests to clarify the nature of the pain and status of disease." All patients are subsequently evaluated by an attending physician (K.M.F. or R.K.P.). The Pain Service physicians follow the patients and supervise analgesic management until discharge or death. The outcome of analgesic therapy is evaluated repeatedly, and modifications to therapy are made when indicated. When deemed appropriate by the patient's primary physicians, the patient is concurrently treated with primary antitumor therapies, such as surgery, radiotherapy, and chemotherapy. Patients assessed as suitable candidates for neurodestructive procedures are referred to the anesthesiology or neurosurgery service, depending on the procedure required. For the current survey, Pain Service physicians pro-

spectively acquired data from 100 consecutive inpatients who were referred over a 14-week period (October 20, 1992, to January 31, 1993). The physician who performed the initial evaluation and provided pain-oriented treatment recorded the following information: patient demographics, reasons for referral to the Pain Service, details of previous analgesic therapy, goals of care, and details of the trials of analgesic therapies implemented until either death or discharge. Patient demographics included age, sex, cancer diagnosis, extent of disease, and pain syndrome. The reasons for referral were selected from a list of 14 options that reflected data from a previous study'' and current impressions of the Pain Service clinicians. Information regarding previous opioid and nonopioid analgesic therapies were derived from the patient history and a review of the patient chart. For the purposes of this survey, opioid therapy was defined as the prescription of a specific opioid (or combination of opioids) by a particular route (or combination of routes) of administration. The conclusion of an opioid therapy was determined by any change in drug or route of administration or the death or discharge of the patient. At the start of each trial of opioid therapy, the Pain Service physician recorded the goal of care (therapeutic intent), the specific opioid and route of administration, and the reasons for the selection of the particular opioid and route of administration. The goal of care was pain relief with preserved cognitive and physical functioning (comfort and function) or pain relief regardless of effect on function (comfort only). Physicians recorded the patient's pain intensity and opioid side effects at the time of referral and at the conclusion of each opioid therapy session. In accordance with current standards of therapy,4813 these data were derived from patient self-report unless the patient was unable to communicate, in which case surrogate evaluation by the pain physician sufficed. Pain intensity was recorded on a 10-point scale, in which 0 indicated no pain and 10 indicated severe pain, and side effects were tabulated. Additional data recorded at the conclusion of each therapy session included the reasons for the change (except when patients were discharged or died) and concurrent medications used to improve analgesia or reduce side effects during the course of the concluded trial. Results

Patient Characteristics The study consisted of 58 men and 42 women, with a median age of 53 years (range, 21-86 years) (Table 1). Ninety-five patients had cancer, four had acquired im-

Opioid Pharmacotherapy for Cancer Pain Management/Cherny et al. Table 1.Patient Characteristics Sex (no.) Male Female Age (yrs) Median Range Tumor type (no.) Breast Colorectal Cervix Lung Sarcoma Lymphoma Pancreas AIDS AML Bladder Myeloma Ovary Other Extent of disease (no.) Metastatic Local/regional Remission Not applicable Pain syndromes (no.) Bone pain Upper GI visceral pain Lumbosacral plexopathy Epidural spinal cord compression Pelvic visceral pain Hepatic capsular distention Chest wall pain Postoperative pain Mucositis Bowel obstruction Brachial plexopathy .Midline retroperitoneal pain Base of skull metastases Herpetic neuralgia Other AIDS: autoimmune deficiencysyndrome; AML: acute myelocyticleukemia; GI:

gastrointestinal.

munodeficiency syndrome, and one had nonmalignant pain. Thirty-two types of cancer were represented. The most prevalent cancers were breast cancer (n = lo), colorectal cancer (n = 9), cervical cancer (n = 7), lung cancer (n = 7), and sarcomas (n = 7). Of the patients with cancer, 92.2% (n = 83) had metastatic disease. One hundred thirty-seven pain syndromes were recorded among the 100 patients. Bone pain (n = 32), upper abdominal visceral pain (n = 13), lumbosacral plexopathy (n = 12), back pain with epidural spinal cord compression (n = 9), pelvic visceral pain (n = 7), hepatic capsu-

1285

lar distention (n = 7), and chest wall pain (n = 7) were the most prevalent syndromes.

Status on Referral One hundred fifty-eight reasons for referral to the Pain Service were recorded (Table 2). The two most common reasons were uncontrolled chronic pain despite analgesic therapy (n = 73) and excessive side effects without adequate analgesia despite analgesic therapy (n = 24). Less common reasons included difficulties in pain assessment and poorly controlled acute pain. At the time of initial evaluation, 77 patients (77%) had a pain intensity of 7-10, and 18 (18%) had a pain intensity of 4-6. After initial evaluation by the Pain Service physicians, the therapeutic intent of analgesic therapy was to provide comfort and function for 87 patients and comfort only for 13 patients. Comfort and function was the therapeutic intent of analgesic therapy for 75 of the patients discharged (92%) and comfort only for 17 patients who died (94%). Ninety-nine of the 100 patients had received prior opioid therapy, which involved a median of two different opioid drugs (range, 1-8 types of drugs) administered by a median of two different routes (range, 1-4 routes). Seventy-seven patients had received nonopioid and adjuvant analgesic agents, of which acet-

Table 2. Reasons for Referral to the Pain Service (N = 100) Reason for referral" Uncontrolled chronic pain despite analgesic therapy Excessive side effects without adequate analgesia Difficulties in assessment Uncontrolled acute pain despite analgesic therapy Change route of opioid administration Concern regarding drug seeking behavior or possibility of addiction Excessive side effects with adequate analgesia Manage preexistent route of drug administration Concern that the patients analgesic regimen involves too many pills Concern that pain is disproportionate for extent of disease Concern that patient using excessive opioid therapy

No. of referrals

73 24 22

12 8

8 5

2

2 1 1

* Physicians were asked to record one or more reasons for referral. The three

reasons that were never cited included uncontrolled pain in a patient not on analgesic therapy, concern that the patient is receiving too many drugs, or concern the the treatment is too expensive.

1286

CANCER October I , 2995, Volume 76,No. 7

I Initial

10&+28 /-

,n=Patients

7

1st Change

+n=Patients

4th Change

Discharged

n=Patients Changed Opioid Therapy

2nd Change

I I

Died

*

Interim tbiereqy

L .\

5th Change

6th Change 7th Change

*\

L

1

\.fl 1

lnltlal theraplom: n-100 Thwaples concludlng in change: n-124 Themplea consludlnp In dbcharge: n=82 Thempi.. concludlng In death: n.18 “-224 Total Number01 th.rap1.S:

Figure 1. The longitudinal course of the 100 patients referred to the Pain Service. Change = change in opioid therapy.

aminophen (n = 47), nonsteroidal anti-inflammatory agents (n = 25), corticosteroids (n = 25), and antidepressants (n = 21) were the most common. The opioids administered at the time of referral included morphine (49%), hydromorphone (33%), oxycodone (14%), fentanyl(1 l%),levorphanol(8%), and methadone (2%).

Course of Therapy After initial evaluation, the Pain Service physicians changed the opioid drug, the route of administration, or both for 58 patients. Of the 42 patients whose regimens were not changed, 17 were discharged, 3 died, and 22 required subsequent changes in either drug or route of administration. Thus, a total of 80 patients required changes in opioid therapy before death or discharge. These 80 patients underwent a total of 182 changes in drug, route of administration, or both (Fig. 1). The specific changes involved a change in route of administration alone (n = 57 [31%]), a change in opioid alone (n = 45 [25%]),deletion of an opioid (n = 36 [20%]),a switch to a new opioid administered by a different route (n = 28 [15%]), addition of a new opioid administered by a new route (n = 15 [8”/.]), addition of a new route of administration without changing the opioid (n = 12 [7%]), discontinuation of one of multiple routes of administration (n = 10 [6?!0]), and the initiation of opioid therapy (n = 1[0.6°/~]).The median duration of hospital stay from time of referral until discharge or death was 9 days (range, 1-52 days). The median duration of a therapy session before a change in drug or route was 3 days (range, 1-30 days). Sixty-four patients required more than one change

in therapy, resulting in a total of 124 changes before discharge or death. The most common reasons for therapy changes subsequent to the initial change were (1)to improve the convenience of treatment in the setting of controlled pain (31.4%), (2) to diminish side effects in the setting of controlled pain (25.0%), (3)to reduce the invasiveness of therapy in the setting of controlled pain (19.3%), and (4) to improve pain control and reduce opioid toxicity simultaneously (17.7%). When opioid toxicity was the reason for a change in therapy, Pain Service physicians changed the opioid drug in 71% of cases and the route in 29%. In response to the need to improve the convenience or reduce the invasiveness of opioid therapy, the physicians changed the route (usually from parenteral to oral or transdermal) in 61% of cases and the opioid (usually to an opioid available in a controlled-release formulation) in 39%. The specific changes in opioid drug are shown in Figure 2. Forty-four patients required one or more changes in opioid before death or discharge, and 20 patients required two or more changes (range, 2-6 changes). Pain Service physicians changed the opioid drug to reduce dose-limiting opioid toxicity in the setting of uncontrolled pain (46%) or pain that was controlled adequately (%%), to improve the convenience of the opioid therapy (25%), or to address a change in pain intensity through a switch between steps 2 and 3 of the World Health Organization’s analgesic ladder (3%). Patients who required a switch in opioid drug had a median stay of 11 days (range, 2-47 days) from time of referral to discharge or death, and patients who required more than one switch had a median stay of 14 days (range 4-47 days). At the time of discharge (n = 82), six patients had experienced sufficient resolution of pain and thus discontinued opioid therapy. Among the patients who continued to require opioid therapy, morphine sulfate was selected more commonly than hydromorphone or fentanyl (39% vs. 23% vs. 17%, respectively). Among the patients who died during the study (n = 18), 16 had continued to require opioid therapy, and hydromorphone was the most commonly used opioid at the time of death. The long half-life opioids, methadone and levorphanol, were tried in 6% and 15%of patients, respectively, and were continued as definitive therapy in 6% and 7% of patients. Several reasons for selection of a specific opioid were cited commonly: (1) the drug had been used previously by the patient and was effective and well tolerated (68.3%), (2) the drug had been used without adverse effects or had not previously been tried (66.5%), and (3) the drug was easy to titrate (50.4%). All of these reasons were cited in the selection of hydromorphone, morphine, and fentanyl. When levorphanol and meth-

Opioid Pharmacotherapy for Cancer Pain Management/Cherny e t al.

Figure 2. Opioids selected.

adone were selected, the absence of previous adverse responses (68%) or a positive previous experience (48%) remained important considerations, but ease of titration was never cited. The less frequently cited reasons for opioid selection, long duration of effect (19.1%), and convenience of the available formulations (18.3%) were most often cited for the selection of controlled-release morphine and transdermal fentanyl. The most common routes of administration at the time of patient referral were oral (67%)and intravenous (33%) (Fig. 3). After an initial evaluation, Pain Service physicians prescribed intravenous opioid to 55% of the patients. Among patients who required further interim changes before definitive therapy at death or discharge, the intravenous route was incorporated either alone or in combination with a second route in 80% of cases. Intraspinal routes were attempted in 12% of patients, but continued until discharge in only 4%. Among the patients discharged with ongoing opioid therapy (n = 76), the routes selected were oral (57%), transdermal (l8%), intravenous (WO), subcutaneous (5%),and intraspinal(4yo). Fifteen of the 16 patients who &ed who had required opioid therapy until death had received the drugs by the intravenous route. Figure 4 lists the reasons cited for the selection of a

Figure 3. Route of opioid administration.

specific route, of which the most commonly cited reasons were convenience (73.6%), noninvasiveness (36.6%), need for rapid effect (36.6%), and impaired gastrointestinal functioning or swallowing or intolerance of oral opioids (23.2%). Whenever a rapid analgesic effect was an important clinical consideration, a parenteral route (i.e., intravenous or subcutaneous) was selected. When patients required very high doses of systemic opioids or when contraindications to oral administration existed, the parenteral routes were again most commonly selected. When patients required rapid relief, neither the oral nor transdermal routes were selected. When patients had impaired gastrointestinal function or swallowing or intolerance to oral opioids, the oral route was never selected. In the absence of these contraindications, the oral route was most commonly selected because of its convenience and noninvasiveness. The intraspinal route was the only route selected for reasons of persistent dose-limiting toxicity from systemic opioid therapy.

Other Drugs Other medications were administered to diminish opioid toxicity or provide co-analgesic effects. These drugs

1288

CANCER October I , 1995, Volume 76, No. 7

Figure 4. Factors influencing selection of route of administration.

were prescribed in combination with 70.1% of opioid therapies, including 79.9% of therapies concluding in change, 57.4% concluding in discharge, and 67.7% concluding in death. Among the patients who required trials of two or more opioids before discharge, use of these agents was reduced from 86% of therapies concluding in change to 54% of therapies at time of discharge. The adjuvant medications used in the surveyed opioid therapies (N = 225) included haloperidol (n = 53 [23%]), corticosteroids (n = 40 [IS”/]), antidepressants (n = 37 [17%]), benzodiazepines (n = 36 [16%]), antiemetics (n = 32 [14%]), psychostimulants (n = 13 [6%]), topical or systemic local anesthetic agents (n = 12 [5%]), nonopioid analgesics (n = 7 [3%])and bisphosphonates (n = 7 [3%]). Antibiotics, baclofen, chlorpromazine, methotrimeprazine, and calcitonin were rarely administered during opioid therapy. Laxatives were prescribed routinely to all patients unless contraindicated.

Clinical Outcome At initial evaluation, 77 of the 100 patients had a pain intensity of 7-10; only 5 patients had rated pain intensity as 0-3. The pain intensity ratings of patients who were continuing to undergo changes in opioid therapy, although improved since presentation, were less satisfactory than those recorded in the interval at discharge or just before death. Among patients who were discharged from the hospital, 77% rated pain intensity in the range of 0-3, and only one (1.2%) had a pain rating of 7-10. Among the 18 patients who died, 17 had a pain intensity rating of 0-3 at time of death. The prevalence of side effects was high (Table 3). Among the patients who were discharged, the prevalence of cognitive impairment, hallucinations, nausea and vomiting, and myoclonus was lower at the time of patient discharge than at referral or at other junctures when opioid therapies were changed, and more patients were free of all recorded side effects at time of

discharge than at other intervals. Among the patients who died, the course of opioid therapy was associated with a particularly high prevalence of somnolence (68.8%), cognitive impairment (43.8%), and hallucinations (37.5%). Discussion

These data, which illustrate the clinical decision-making process of pain specialists during treatment of selected inpatients with cancer pain, contribute to an understanding of specific therapeutic strategies, including (1)the influence of goals of care on decision making, (2) selection of opioid drugs, (3) indications for changing opioid drugs and the frequency with which this strategy

Table 3. Prevalence of Physician-Recorded Opioid Side Effects Recorded Before Initial Pain Service Treatment, at the Time of Change of Interim Trials of Therapy, and Definitively at Time of Discharge or JustBefore Death

Side effects Nil Somnolence Constipa tion Cognitive impairment Hallucinations Nausea/vomiting Myoclonus Mood disturbance Seizure Mood disorder Urinary retention Dry mouth Itch Local irritation Dizziness Respiratory depression

Prior

Change

Discharge

Before death

N=lOO

N=124

N=76

N=16

21 38 34 25 11

20.1 35.4 21.7 33.1 14.5

11 9 7 2 1

1 1 0 0 0 0

30.2 35.5 39.4 13.1 3.9

31.2 68.8 12.5 43.8 37.5

12.9

7.8

12.5

8.8 13.7 1.6 0 0.8 0 0.8 0 2.4 0.8

2.6 6.5 0 0 0 1.3 0 0 0 0

0 6.3 0 0 0 0 0 0 0 6.3

Opioid Pharmacotherapy for Cancer Pain Management/Cherny ef al.

is used, and (4) selection of route of administration. Although the pain intensity and adverse effects data were typically derived from patient self-report in accordance with routine practice, this methodology was not designed to provide a valid and detailed assessment of subjective outcomes. Despite this limitation, the data suggest that most patients had poorly controlled pain at the time of referral, but ultimately experienced more satisfactory control at the time of discharge or death. The process by which this was achieved began with a careful pain assessment that evaluated the pain characteristics and syndrome, the putative mechanisms sustaining the pain, and the effect of the pain on function and psychologic well-being. In addition, the assessment defined the nature and extent of the underlying disease, identified the goals of care, and noted prior analgesic therapies and their outcomes. It has been suggested that the evaluation of the goals of care provides an essential context for therapeutic decision m a l ~ n g . When '~ patients prioritize optimal comfort and function equally, the therapeutic intent is to achieve an adequate degree of relief without compromising cognitive and physical function. When comfort is the overriding goal of care, there is a willingness to use whatever analgesic therapy is necessary to achieve relief, even if function is diminished in the process. For the vast majority of patients who were ultimately discharged, the goals of care were comfort and function. This was reflected in the substantial effort to optimize the balance between relief and side effects through dose adjustment, sequential trials of opioid drugs, co-administration of adjuvant medications, and the use of spinal analgesic techniques. In contrast, comfort alone was the overriding goal of care for 17 of the 18 patients who died. Among these patients, somnolence and cognitive impairment were more prevalent and were rarely treated; sedating adjuvant drugs were co-administered frequently to reduce agitation or treat distress. The strategy for drug selection implemented by the Pain Service physicians incorporates the following principles: The patient's analgesic history is reviewed, with particular attention paid to the presence of adverse effects. If the current opioid is well tolerated, this drug is continued unless difficulties are anticipated in dose titration or in administration of the required dose in a convenient formulation. In these situations, an alternative opioid is considered. If the current opioid is producing dose-limiting toxicity, an alternative opioid is usually selected. When choosing an alternative opioid, the physicians choose the drug that best combines ease of titration and convenience in dosing among those opioid drugs that had not previously caused uncontrolled adverse effects. The Pain Service physicians most commonly se-

1289

lected short half-life opioid agonists, specifically morphine, hydromorphone, and fentanyl. All other factors being equal, these drugs were preferred because the shorter period required to approach steady-state plasma concentrations simplifies dose titration. The range of available formulations influenced the selection among these drugs. This is illustrated by the low frequency with which oxycodone was selected in the management of severe pain. Although oxycodone is a versatile opioid agonist, the doses required by patients with severe pain could not be administered conveniently using the oral formulations available, and no parenteral formulation is available. For ambulatory patients who were able to tolerate oral opioids, morphine sulfate was most often used, in part because the range of available formulations (including the controlled-release form) was most flexible and potentially most convenient. Future opioid selection is likely to be influenced by the availability of new formulations, including sustained-release formulations of oxycodone and hydr~morphone,'~,'~ a new type of controlled-release m ~ r p h i n e , ' ~ -and ' ~ a transmucosal preparation of fentanyl.20,21 In the current survey, 44 patients (44%) required trials of two or more systemically administered opioid drugs, and 20 required sequential trials of three or more opioids until discharge (n = 18) or death (n = 2). This group of 20 patients underwent a total of 52 changes in opioid selection (median, 2 changes; range, 2-6 changes) for the reasons outlined above. Pain Service physicians switched opioids when patients developed dose-limiting toxic reactions in the setting of either uncontrolled (43%) or controlled pain (27%) or when a more convenient opioid regimen was desirable (25%). The most common dose-limiting side effects were cognitive impairment, hallucinations, myoclonus, and nausea. Many patients experienced an improvement in the balance between adequate analgesia and adverse effects when an opioid that had been producing intolerable side effects was replaced by another. The benefits of this strategy are evidenced by the increase in the proportion of patients with no adverse effects (from 18% to 41%) and by the diminished requirement for co-analgesics and therapies to manage adverse effects (from 86% to 54%) among the patients who received interim therapies and were ultimately discharged (N = 37). The potential value of sequential trials of different opioid drugs has been suggested clinically by the large intraindividual differences in responses to various opioids. Small series of patients who achieved a good response to an opioid after developing dose-limiting adverse effects from other opioids have been A retrospective evaluation of 110 palliative care inpatients revealed that 32 of the 44 patients who underwent a

1290

CANCER October I , 2995, Volume 76,No. 7

change in opioids due to delirium improved thereafter.25The data from the current survey suggest that this strategy is frequently required and that physicians should have a therapeutic repertoire of at least three opioid drugs used in the management of severe pain (i.e., morphine, hydromorphone, fentanyl, oxycodone, oxymorphone, methadone, and levorphanol). To use the technique of sequential opioid trials safely, the clinician must be familiar with the use of equianalgesic dosing data to calculate appropriate starting doses.I4 For patients who have a limited life expectancy, the speed with which systemic opioid trials can be accomplished may be an important consideration in selecting between further trials of alternative opioid drugs and the taking of other approaches, such as spinal drugs or neurolysis. This survey demonstrated that, in an inpatient setting, sequential opioid trials can be accomplished with a median duration of only 3 days per drug. In selecting the route of opioid administration, we found the severity of the pain and the rapidity with which relief was required to be critical considerations. Opioid administration by a parenteral route yields a shorter time to peak analgesic effect compared with oral or transdermal routes, thus facilitating more rapid dose adjustment. When patients experienced severe or excruciating pain, Pain Service physicians generally selected a parenteral route to facilitate rapid titration of the opioid dose. In a previous survey of intravenous opioid infusion, this was the overwhelming reason for the selection of this route.26 When stable pain control can be achieved at a dose that can be administered by a noninvasive route, patients are generally switched to such a route before discharge. Most patients in this survey were discharged using the oral (53%)or transdermal (18%)route of administration. The 23% of discharged patients who continued to receive parenteral opioid therapy were those who had a contraindication to the oral route, required very high doses of systemic opioids, or needed rapid relief from severe breakthrough pain. The common use of the intravenous route among these patients reflected the high prevalence of infusion ports previously inserted for the administration of chemotherapy. Subcutaneous infusions were used for patients without preexisting intravenous ports, unless the selected opioid was irritating or the rate of infusion (cc/ hour) was poorly tolerated. Transdermal opioid administration with the Duragesic fentanyl patch (Jonson Pharmaceutical, Beerse, Belgium) was used by 18% (n = 14) of patients who were receiving opioid therapy at discharge. Because fentanyl is the only drug available by this route and the transdermal formulation is the only noninvasive formulation routinely available for this drug, the decision

to use this drug and route were influenced by drug selection factors and concerns directly related to route. Of these 14 patients, 6 had undergone a prior trial of intravenous fentanyl and had benefitted from this opioid. In the United States, the availability of transdermal fentanyl has justified trials of intravenous fentanyl among patients with chronic pain, because a favorable outcome can now be followed by a convenient long term therapy. Among the patients discharged using transderma1 fentanyl, supplemental doses for breakthrough pain were provided by another opioid and route for six of these patients or as parenteral fentanyl administered through a patient-controlled analgesia device for two patients. In the current survey, opioid administration by either the epidural or intrathecal route was attempted in 12% but continued in only 4% of patients. In the absence of randomized trials that compare the various intraspinal techniques with other analgesic approaches, the indications for the selection of a spinal route remain empirical. Some authorities prefer these techniques to parenteral routes for patients who are unable to tolerate oral ~pioids,’~ and others advocate a trial of an intraspinal approach rather than sequential opioid trials in the setting of dose-limiting toxicity. In our institution, persistent dose-limiting toxicity from systemic opioid therapy was the usual indication for the selection of this route. Hogan et al., who described similar guidelines for patient selection, reported that only 16 of 1205 cancer patients with pain required use of these routes.” Although simplicity of prescription is preferred, some patients required drug combinations to improve the balance between the analgesic and adverse effects of systemic opioid therapy by diminishing opioid toxicity or providing co-analgesic effects. The diverse characteristics of medications used by the Pain Service physicians reflects the growing pharmacotherapeutic complexity of this evolving field.29 Although controlled clinical trials have begun to provide a scientificallyvalid assessment of many of these agent^,^'-^^ clinical use continues to be largely empirical. The drugs most commonly co-administered by Pain Service physicians were used to treat delirium and neuropathic pain. Delirium is common among dying patient^*^-^^ and is a commonly observed dose-limiting adverse effect of opioid medicat i o n ~Given . ~ ~ that opioids were frequently titrated to dose-limiting toxicity in this medically ill inpatient population with difficult pain problems, it is not surprising that delirium was a common clinical problem and that haloperidol, a neuroleptic used to manage delirium, was the most commonly prescribed adjuvant. Haloperidol was added to almost one in three therapies that concluded in either change (usually due to dose-limiting toxicity) or death. Corticosteroids, antidepressants,

Opioid Pharmacotherapy for Cancer Pain Management/Cherny et al.

1291

Table 4. Principles in the Selection of Opioid Drug and Route of Administration Selection of opioid drugs Evaluate the pain and review the patient’s analgesic history with particular attention to the presence of adverse effects from present or previous opioid therapy. If the patient is opioid naive and has strong pain, morphine sulphate is generally recommended because the range of available formulations and widespread physician familiarity. If the patient is already using an opioid that is well tolerated, it should be continued with appropriate dose titration unless difficulties are anticipated in dose titration or in administration of the required dose in a convenient formulation. A switch to an alternative opioid is considered in the following circumstances: if the patient develops dose-limiting toxicity that precludes adequate relief of pain without excessivc side effects if a specific formulation, not available with the current drug, is either needed or may substantially improve the convenience of opioid administration. If a switch in opioid is indicated the following considerations are salient: short half-life opioid agonists, specifically morphine, hydromorphone, fentanyl, oxycodone, and oxymorphone are usually preferred over long half life drugs such as methadone and levorphanol because of easier dose titration and reduced risk of delayed toxicity from dose accumulation. some agents are only available in a limited range of formulations which may influence drug selection: oxymorphone is available only as a rectal suppository or for injection, fentanyl is only available for transdermal or parenteral administration, oxycodone is only available for oral administration. some patients will require sequential trials of several different opioids before a drug that is effectice and well tolerated is identified. If systemically administered opioids are persistently unable to achieve adequate relief without excessive adverse effects despite switching opioids, consideration is given to spinal opioids or other regional anesthetic or neuroablative therapies. Selection of route of administration When pain is severe and the rapid relief is required, parenteral administration is preferred because the shorter time to peak analgesic effect facilitates more rapid dose adjustment. When stable pain control is achieved or the need for pain relief is not immediate, the oral route is recommended When patients are unable to tolerate oral medication rectal, transdermal and parenteral routes are considered. If the patient requires very high doses of systemic opioids, or rapid relief from severe breakthrough pain, then the parenteral routes are preferred. If the patient had preexisting IV access this route should be used, otherwise doses can usually be administered subcutaneously. If the patient does not require very high doses of systemic opioids or rapid relief from severe breakthrough pain, then either the transdermal, rectal, or sublingual routes can also be considered. Selecting between these routes is also influenced by considerations regarding drug selection, cost, and patient preference. Spinal routes of opioid administration are considered when the systemic administration of opioids therapy is persistently unable to achieve adeauate relief without excessive adverse effects. iv: intravenous.

and systemic local anesthetic agents, such as lidocaine or mexiletine, were commonly administered in the management of neuropathic pain syndromes. Dexamethasone was used routinely in the treatment of epidural spinal cord compression and commonly in the management of pain associated with the compression of other neural structures and severe bone pain.51Psychostimulants, such as methylphenidate, dextroamphetamine, and pemoline were used to treat opioid-induced somnolence or cognitive impairment. This survey illustrates the application of principles described in current guidelines for cancer pain management.l.3,4,7Specifically, the data depict an accepted approach to the selection of opioid drugs and routes of administration (Table 4). Because these data were accrued in the treatment of highly selected patients in a unique inpatient setting, the frequency data cannot be extrapolated to outpatient care of unselected patients. Nonetheless, the frequency with which experienced clinicians make treatment changes in response to the

changing clinical situation highlights the degree of monitoring and clinical flexibility that may be required to manage cancer pain in accordance with contemporary standards. References 1. Foley KM. The treatment of cancer pain. N EnglJMed 1985:313: 84-95. 2. Twycross RC, Lack SA. Symptom control in far-advanced cancer: pain reliel. London: Pitman, 1984. 3. Chemy NI, I’ortenoy RK. Cancer pain management: current strategy. Cancer 1993;72(Suppl):3393-415, 4. American Pain Society. Principles of analgesic use in the treatment of acute pain and chronic cancer pain: a concise guide to medical practice. 3rd ed. Skokie (IL): American Pain Society, 1992. 5. World Health Organization. Cancer pain relief. Geneva: World Health Organization, 1986. 6. World Health Organization. Cancer pain relief and palliative care. Geneva: World Health Organization, 1990. 7. Agency for Health Care Policy and Research. Acute Pain Man-

1292

8.

9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

CANCER October 2, 2995, Volume 76,No. 7

agement Panel: management of cancer pain, clinical practice. Guideline No. 9. Washington, DC: U.S. Department of Health and Human Services, 1994. Von Roenn JH, Cleeland CS, Gonin R, Hatfield A, Pandya KJ. Physician's attitudes and practice in cancer pain management: a survey from the Eastern Cooperative Oncology Group. A n n InternMed 1993;119:121-6. Cleeland CS, Gonin R, Hatfield A, Edmonson MD, Blum RH, Stewart JA, et al. Pain and its treatment in outpatients with metastatic cancer. N Engl l M e d 1994;330:592-6. Grecco PJ, Eisengerg JM. Changing physician's practices. N Engl J M e d 1993;329:1271-4. Gonzales GR, Elliot KJ, Portenoy RK, Foley KM. The impact of a comprehensive evaluation in the management of cancer pain. Pain 1991;47:141-4. Moulin DE, Foley KM. A review of a hospital based pain service. In: Foley KM, Bonica JJ, Ventafrida V, editors. Second lnternational Congress on Cancer Pain: advances in pain research and therapy. vol. 16. New York: Raven Press, 1990:413-28. American Pain Society Committee of Quality Assurance Standards. American Pain Society quality assurance standards for the relief of acute pain and cancer pain. In: Bond MR, Charlton JE, Woolf CJ, editors. Proceedings of the Sixth World Congress on Pain, Pain Research and Clinical Management. vol 4. Amsterdam: Elsevier, 1991:185-90. Cherny NI, Portenoy RK. Practical issues in the management of cancer pain. In: Wall PD, Melzack R, editors. Textbook of pain. 3rd ed. Edinburgh: Churchill Livingstone, 1994:1437-67. Bruera E, Sloan P, Scott J, Mount B. Slow release (SR) versus intermediate release (IR) hydromorphone (HM) in patients with cancer pain: a controlled, crossover trial. In: Seventh World Congress on Pain: Congressional Abstracts. Seattle: IASP Publications, 1993: Abstract 899. Hays H, Hagen N, Thirlwell M, Dhaliwal H, Babul N, Harsanyi Z, et al. Comparative clinical efficacy and safety of immediate release and controlled release hydromorphone for chronic severe cancer pain. Cancer 1994; 74:1808-16. Plummer JL, Cherry D, Gourlay G, Onley MM, Hooper MB. A comparison of KapanoIB (a new sustained release morphine formulation) and MST Continus@ and morphine solution in cancer patients: pharmacodynamic aspects. In: Seventh World Congress on Pain: Congressional Abstracts. Seattle: IASP Publications, 1993: Abstract 1000. Gourlay G, Plummer DA, Cherry D, Onley MM, Fugere F. A comparison of KapanolO (a new sustained release morphine formulation) and MST Continus@and morphine solution in cancer patients: pharmacokinetic aspects. In: Seventh World Congress on Pain: Congressional Abstracts. Seattle: IASP Publications, 1993: Abstract 998. Cherry DA, Gourlay G, Plummer JL, Onley MM, Russell A, Beaumont AC. A comparison of KapanolR (a new sustained release morphine formulation) and MST Continus@ and morphine solution in cancer patients: morphine metabolite profiles and renal function. In: Seventh World Congress on Pain: Congressional Abstracts. Seattle: IASP Publications, 1993: Abstract 999. Stanley TH, Hague B, Mock DL, Streisand JB, Bubbers S, Dzelzkalns RR, et al. Oral transmucosal fentanyl citrate (lollipop) premedications in human volunteers. Anesth Anulg 1989;69:21-7. Fine PG, Marcus M, DeBoer A], Van der Oord B. An open label study of oral transmucosal fentanyl citrate (OTFC) for the treatment of breakthrough cancer pain. Pain 1991;45:149-55. MacDonald N, Der L, Allan 5, Champion P. Opioid hyperexcit-

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

36.

37.

38.

39.

40.

41.

ability: the application of alternate opioid therapy. Pain 1993;53: 353-5. Crews JC, Sweeney NJ, Denson DD. Clinical efficacy of methadone in patients refractory to other mu opioid receptor agonist analgesics for management of terminal cancer pain: case presentations and discussion of incomplete cross-tolerance among opioid agonist analgesics. Cancer 1993; 722266-72. Galer BS, Coyle N, Pasternak GW, Portenoy RK. Individual variability in the response to dlfferent opioids: report of five cases. Pain 1992;49:87-91. de Stoutz ND, Bruera E, Suarez-Almazor M. Opiate rotation (OR) for toxicity reduction in terminal cancer patients. In: Seventh World Congress on Pain: Congressional Abstracts. Seattle: lASP Publications, 1993: Abstract 894. Portenoy RK, Moulin DE, Rogers A, Inturrisi CE, Foley KM. I.V. infusion of opioids for cancer pain: clinical review and guidelines for use. Cancer Treat Rep 1986;70:575-81. Plummer JL, Cherry DA, Cousins MJ, Gourlay GK, Onley M, Evans KHA. Long term spinal administration of morphine in cancer and non-cancer pain: a retrospective study. Pain 1991; 44: 212-20. Hogan Q, Haddox JD, Abram S, Weissman D, Taylor ML, Janjan N. Epidural opiates and local anesthetics for the management of cancer pain. Pain 1991;46:271-9. Portenoy RK. Adjuvant analgesics in pain management. In: Doyle D, Hanks GW, MacDonald N, editors. Oxford textbook of palliative medicine. Oxford: Oxford University Press, 1993:187203. Della Cuna GR, Pellegrini A, Piazzi M. Effect of methylprednisolone sodium succinate on quality of life in preterminal cancer patients: a placebo-control multicenter study. Eur J Cancer Clin Oncol 1989; 29: 1817-21. Tannock I, Gospodarowicz M, Meakin W, Panzarella T, Stewart L, Rider W. Treatment of metastatic prostatic cancer with lowdose prednisone: evaluation of pain and quality of life as pragmatic indices of response. I Clin Oncol 1989; 7:590-7. Greenberg HS, Kim J, Posner JB. Epidural spinal cord compression from metastatic tumor: results with a new treatment protocol. A n n Neurol 1980; 8:361-6. Watson CPN, Evans RJ, Watt VR. Posttherapeutic neuralgia and topical capsaicin. Pain 1988; 33:333-40. The Capsaicin Study Group. Treatment of painful diabetic neuropathy with topical capsaicin: a multicenter, double-blind, vehicle-controlled study. Arch Intern Med 1991; 1512225-9. Scheffler NM, Sheitel PL, Lipton MN. Treatment of painful diabetic neuropathy with capsaicin 0.075%. J A m Podiatr Med Assoc 1991;81:288-93. Tandan R, Lewis GA, Krusinski PB, Badger GB, Fries TJ. Topical capsaicin in painful diabetic neuropathy: controlled study with long-term follow-up. Diabetes Cure 1992; 15:8-14. Watson CPN, Evans RJ, Watt VR. The post-mastectomy pain syndrome and the effect of topical capsaicin. Pain 1989;38:17786. Coleman RE, Woll PJ, Miles M, Scrivener W, Rubens RD. Treatment of bone metastases with (3 amino-I-hydroxypropy1idene)1, 1-bisphosphonate (APD). B r l Cancer 1988;58:621-5. Thiebaud D, Leyvraz S, von Fliedner V, Perey L, Cornu P, Thiebaud S, et al. Treatment of bone metastases from breast cancer and myeloma with pamidronate. EurJ Cancer 1991; 27:37-41. Morton AR, Cantrill JA, Pillai GV, McMahon A, Anderson DC, Howell A. Sclerosis of lytic bony metastases after aminohydroxypropylidene bisphosphonate (APD) in patients with breast cancer. B r l M e d 1988;297:772-3. van Holten Verzantvoort ATM, Zwinderman AH, Aaronson

Opioid Pharmacotherapy for Cancer Pain Management/Cherny et al.

42.

43.

44.

45.

NK, Hermans J, van Emmerik B, van Dam F, et al. The effect of supportive pamidronate treatment on aspects of quality of life of patients with advanced breast cancer. EurJCancer 1991;27:5449. Clarke NW, McClure J, George NJR. Clinical and metabolic effects of disodium pamidronate in metastatic prostate cancer. In: Bijvoet OLM, Lipton A, editors. Osteoclast inhibition in the management of malignancy-related pain. Lewiston, NY: Hogrefe & Huber, 1991:54-64. Ernst DS, MacDonald RN, Paterson AHG, Jensen J, Brasher P, Bruera E. A double blind, crossover trial of I.V. clodronate in metastatic bone pain. J Pain Sympt Management 1992; 7:4-11. Porter AT, McEwan AJ, Powe JE, Reid R, McGowan DG, Lukka H, et at. Results of a randomized phase-IIl trial to evaluate the efficacy of strontium-89 adjuvant to local field external beam irradiation in the management of endocrine resistant metastatic prostate cancer. Int JRadiat OncolBio2 Phys 1993;25:805-13. Lewington VJ, McEwan AJ, Ackery DM, Bayly RJ, Keeling DH, Macleod I'M, et al. A prospective, randomised double-blind

46. 47.

48.

49. 50.

51.

1293

crossover study to examine the efficacy of strontium-89 in pain palliation in patients with advanced prostate cancer metastatic to bone. EurJCancer 1991;27:954-58. Massie MJ, Holland JC, Glass E. Delirium in terminally ill cancer patients. Am I Psychiatry 1983; 140:1048-50. Ventafridda V, Ripamonti C, De Conno F, Tamburini M, Cassileth BR. Symptom prevalence and control during cancer patients' last days of life. ] Palliate Care 1990;6:7-ll. Bruera E, Chadwick S, Weinlick A, MacDonald N. Delirium and severe sedation in patients with terminal cancer. Curr TherapeuticRes 1987;71:787-8. Fainsinger R, Young C. Cognitive failure in a terminally ill patient. J Pain Sympt Management 1991;6:492-4. Bruera E, Macmillan K, Hanson J, MacDonald RN. The cognitive effects of the administration of narcotic analgesics in patients with cancer pain. Pain 1989;39:13-6. Ettinger AB, Portenoy RK. The use of corticosteroids in the treatment of symptoms associated with cancer. ] Pain Sympt Management 1988; 3:99-103.