- Email: [email protected]
Management of cancer pain with transdermal fentanyl: Phase IV trial, University of Iowa
El. 7No. 3 (St&&)Aprit1992
of Pain and S$mphmManagement 2% 3oumal
Management of Cancer Pain with Transdermal Fentanyl: Phase IV Trial, University of Iowa Timothy J. Maves, MD, and Winston A. Barcellos, MD Departmentof Ane&iesk.a,Lliiversityof Iowa GI&ge
ofMedicine,Iowa City, Iowa
AA multicenterstudy was conducted to determinethe patient andphysicianacc@abiIi& oftransahnal fentq_vl in the management of cancer-related pain. In this s&y, IO cancer patients at the Universi4yof Iowa received transdermalfmtunyl afler discontinuingtheirprior optid anakesic; ? p&?&s completed questionnaires before and at 2 and 4 wkfolhxuing transdermalf~tanyl application. ?here was no s&n$&nt d@rence in vimaL analogue scale scoresforpain or mood. Erbalpain descr$tor scores improved at 2 wk (P c .05). lltere was a nons&$cant tenden~ toward increased d$ression and nauseq’ however, patients spent less t&e thinking about their ii&ss andfet their cancer was Lessdt&uptiueto their closest j%uWrelatiues. Constipation, appetite, drowsiness, and concentration were not stahSal& d@rent. Pati reported improved sleep habits at 2 wk (P < .05) and tended to require less help z&h eating, drRtrbrg,washing, and using the bathroom. Allpa&& completi~ the sttiy chose to continue transdermal fentanylfm their cancer pain management. In summcuy, these data demonstrate the anaigesic efiq of the transdermalfentanyl system and suggest that some patients with cancer-relaied pain could benejtjom its ue. J Pain Symptom Manage 1992;7:S58+62.
&YWd
Pain, titeraH, cancer, narcotics
Introduction Appropriate management of cancer-related pain requires 1) adequate knowledge of the pain mechanism(s) involved, 2) inclusive pain assessment,’ 3) identification of barriers to adequate pain control,* and 4) a multifaceted approach to pain therapy. For some patients, current therapeutic options cause signilicant adverse effects, are inconvenient, provide inadequate analgesia, or are expensive. Transdermal fentanyl is a novel addition to the armamentarium against cancer-related pain. Small surveys suggest that some patients treated with Ada& reprintrequeststo: Timothy J. Maves, MD, Pain Clinic, Department of Anesthesia, University of Iowa College of Medicine, Iowa City, IA 52242. @ U.S. Cancer Pain Relief Committee, 1992 Published by Elsevier, New York, New York
transdermal feutanyl will experience relatively fewer side effects3 and have steady-state fentanyl blood levels that fluctuate less than with opioids delivered by many other routes.4 The transdermal route also may be convenient for patients and cost competitive with other noninvasive therapeutic options. As part of a multicenter study, we examined the effect of transdermal fentanyl on the quality of lie of cancer patients.
Methods The experimental protocol was reviewed and approved by the University of Iowa Hospitals and Clinics Human Subjects Review Committee. A total of 10 cancer patients received transdermal fentanyl. Inclusion criteria included a histologically confirmed malignancy, pain related to the cancer
om-3924/92/85.00
til. 7.h.
3 (Sub@) April 1992
I’huse IV Trial ojTransderma1 Fentanyl
and/or therapy, pain necessitating the use ofopioid analgesics, age of 18 yr or older, and ability to communicate effectively with study personnel. Following informed consent, each patient completed 1) visual analogue scale (MS) for pain intensity and mood, 2) verbal pain descriptor (VPD, series of words describing the intensity of pain, with each word assigned a numerical score), 3) Functional Living Index--Cancer (FLJC), and 4) a symptom questionnaire (SQ). History, physical examination, and appropriate laboratory studies were obtained for each patient. Patients previously requiring opioid analgesics were converted to an intramuscular (IMJ morphine equivalent dose using the equianalgesic potency conversions reported by F01ey.~ The analgesic activity ratio of 10 mg IM morphine every 4 hr to 100 pg intravenous (IV) fentanyl every hour was used to derive the initial transdermal fentanyl dose. Prior opioid analgesics were discontinued, and all sedative-agent do.sages were halved when transdennal fentanyl was applied. Opioid-naive patients .were started on 25 gg/hr transdermal fentanyl. Breakthrough pain was treated with short-acting oral opioids (10 mg
s59
morphine or %mg hy~omo~hone) every 20 min as needed for pain control. TransdermaS fentanyl dosage was increased every 2-3 days as indicated by the daily dosage requirements for breakthrough medication. Patients were followed by phone OF personal interview every 1-3 days. Any side effects were treated promptly with appropriate agents or interventions. VAS, VPD, FLIC, and SQ were completed by each patient 2 and 4 wk following initial transdermal fentanyl application. The questionnaire data were analyzed by calculating the mean f SEM for each question. Data obtained prior to treatment with transdermal fentanyl were compared to data at 2 and 4 wk using paired Student’s t test.
Patient profiles are illustrated in Table I; 7 patients completed the study, 1 patient died 2 wk into the study, a second developed septic shock and transdermal fentanyl was diicontinued, and the third patient was withdrawn for increase; nausea and vomiting during chemotherapy. Tabb I
Patient Profile Initial tKillSdcrmal
Dose at
dose
study
fentanyl end of Cancer
Age (yr)
Sex
type
Prior TX
Prior analgesic
(&hr)
1
45
F
XRT
2
60
M
Lun_g (WDNW Lung
3
51
M
Luns
XRT
4
68
M
‘SC) Lung
XRT
Controlled-release morphine Controlled-rclcasc morph& Controlled-release morphine Controlled-rclcasc
5
70
M
(PDNSC) Sq. Ceil BOT
XRT
Controlled-release
Patient
6
65
F
7
54
F
Breast (Infil. Ductal) Cervical
8
64
M
Prostate
9
52
F
10
48
M
Colon (Adeno) Non-Hodgkin’s lymphoma
aDicd
Chcmo
Chcmo Surgical XRT, Chemo Surgical Chcmo, Rad Surgical XRT$ Hormonal Surgical XRT, Chcmo Surgical Chcmo
at 2 wk elevated liver hnction tests. bDiscontinued at 2 wk titer developing G-sepsis. CDiscontinued at 2 wk secondary to naura.
morphine
(&hr)
Pain types
50
1ooa
25
75
Bone mets, leftbra&al plexopathy Bone mets
75
500b
25
50
25
125
Tumor related
50
100
Left bra&al
50
125
Tumor related
25
150
Bone mets
100
150
Tumor related
Bone mets Bone and liver mets
l7lOl$liIlC
Controlled-release morphine Controlled-release morphine Controlled-release morphine Hydromorphone Controlled-rclcasc morphine
50
5oc
Tumor relp!
plcxopathy
RL. 7No. 3 (Suppl.)April 1992
Mwes and Barcelhs
S60
100 z
E s D .-is P
a0 60 40 20 0
a
I pr9-
2-week
4-week
pre-
P-week
4-week
a
b
b
-
pre-
2-week
4-week
pre-
2-week
4-week
pre-
2-week
4-weetc
Fig. 1. Visual analog scale (mm) for (a) pain (0 = least
possiblepain, 100 = worst possible pain) and (b) mood (0 = best mood, 100 = wont mood) pre-, 2 wk, and 4 wk after transdermal fentanyl application, recorded as mean f SEM.
There was no significant difference in VAS for pain or mood at 2 or 4 wk compared with pretreatment values (Figure 1). Transdermal fentanyl improved analgesia, as evidenced by lower VPD score at 2 wk (PC .05), but VPD score at 4 wk was not statistically different (Figure 2a). Analysis of the FLIC revealed a nonsignificant tendency toward increased depression and nausea (Figure 2b and c, respectively). Although not signifrcandy different, patients spent less time thinking about their illness and felt their cancer was less disruptive to their closest friends/relatives (Figure 3a and b, respectively). There was no significant diirence in constipation, appetite, drowsiness, or concentration when analyzing the SQ Patients did report improved sleep habits at 2 wk (P c .05), but the difference was not significant at 4 wk (Figure 4a). Patients tended to require less help with eating, dressing, washing, and using the bathroom (Figure 4b). All patients completing the study chose to continue transdermal fentanyl for their cancer pain management (Figure 4c).
6 85 $4 !I In 3 3 c 2 1 0 C
Fig. 2. Functional living index for cancer scores: (a) pain description scale (0 = no pain, 7 = excruciating), @) depression scale (0 = never, 7 = continually), and (c) nausea scale (0 = none, 7 = a great deal) pre-, 2 wk, and 4 wk after transdermal fentanyl application, recorded as mean f SEM (*P < .05).
Discussion Steady-state fentanyl blood levels resulting in increased analgesia and uninterrupted sleep could potentially have a great impact on quality of life for cancer patients. Although the data did not reveal marked improvement in functional abilities, many spouses expressed that their wives/husbands were more alert and interactive while using transdermal fentanyl. All patients completing the study expressed a desire to continue transdermal fentanyl for their pain management. Bypassing the gastrointestinal tract during opioid
S61
2-week
$-week
P-week
Q-week
b pre-
pre-
2-week
B-week
pm-
P-week
$-week
Fig. 3. Functionalliving index for cancer scores:(a) time spent thinking ahout their illness (0 = never, 7 = constantly),and (9) cancer-relateddisruptionto relatives (0 = no disruption,7 = totallydisruptive)pre-, 2 wk, and 4 wk after transdermalfentanylapplication,recorded as mean * SEM. administration could contribute to a smaller incidence of nausea and constipation.‘j Iftrue, the small number of patients in our study may account for the lack of improvement in these symptoms. Respiratory depression, pruritus, confusion, and allergic reaction did not occur in our patients during transdermal fentanyl application. Of the 10 patients, 9 required at least a 500/o increase in transdermal fentanyl dose during the study. The dose escalation may have resulted from inadequate analgesia prior to the study, relatively low initial fentanyl dose secondary to a conservative conversion ratio (6:1, PO to IM morphine), progression of disease, and/or tolerance. Opioid tolerance may depend on efficacy of the specific opioid (a high-efficacy agonist may down-regulate fewer receptors and produce less tolerance)7 and method of delivery (it has been suggested that continuous IV or intrathecal routes develop tolerance faster than intermittent oral dosing, but this has not been confiied).8 Tolerance to transdermal fentanyl should be addressed in future studies. In summary,
transder.mal
fentanyl
offers an
G
P-week
4-week
Kg. 4. Symptom questioriaire: (a)frequency at which pain preventeda good night’ssleep(0 = none of the time, 5 = all of the time); (b) frequency at which patient required help eating, dressing,washing, and using the bathroom(0 = none of the time, 5 = all of the time)pre-, 2 wk, and 4 wk tier transdermalfentanyl application, recorded as mean +zSEM CP < .05); (c) percentage of patientswho would use transdermalfentanylagain.
efficacious and convenient alternative to conventional opioid therapy. Data from this open-label trial suggest that some patients could benefit when transdermal fentanyl is utilized in the management of cancer-related pain.
We thank Michael Burcham and Linda Hendricks for preparation of the graphic and Peggy Miller for assistance with patient evaluation.
862
Maves and Bmceh
Refeemes 1. F&y KM. Assessment of pain in patients with cancer. In: Swerdlow M, Ventafiidda V, eds. Cancer pain. Norwell, MA: MTP Press, 1987~37-57. 2. Ieti DN, Cleeland CS, Dar R Public attitudes toward cancer pain. Cancer 1985;56:2337-2339. 3. Simmons MA, Payne R, Richenbacher J, Moran K, Southam MA. ‘ITS (fentanyl) in the management of pain in patients with cancer. Proc Am Sot Clin Oncology 1989;8:234. 4. Miser AW, Narang PK, Dothage JA, Young RC, Smdelar W, Miser JS. Transdermal fcntanyl for pain
Vol. 7flo. 3 (Supp f.) Afnif 1992
control in patients with cancer. Pain 1989;37: 15-2 1. 5. Foley KM. The treatment of cancer pain. N Engl J Med 1985;313:84-95. 6. Ventaftidda V, Spoldi E, Caraceni A, Tamburini M, DeConno F. The importance of continuous subcutaneous morphine admiistration for cancer pain control. Pain Chn 1986; 1:47-55. 7. Sosnowski M, Yaksh TL. Differential cross-tolerance between intrathecal morphine and sufentanil in the rat. Anesthesiology 1990;73:1141-1147. 8. JaffeJH. Drug addiction and drug abuse. In: Goodman GA, Goodman Is, Rail TW, Murad F, eds. The pharmacological basis of therapeutics. New York Pergamon Press, 1990:533.
of Pain and S$mphmManagement 2% 3oumal
Management of Cancer Pain with Transdermal Fentanyl: Phase IV Trial, University of Iowa Timothy J. Maves, MD, and Winston A. Barcellos, MD Departmentof Ane&iesk.a,Lliiversityof Iowa GI&ge
ofMedicine,Iowa City, Iowa
AA multicenterstudy was conducted to determinethe patient andphysicianacc@abiIi& oftransahnal fentq_vl in the management of cancer-related pain. In this s&y, IO cancer patients at the Universi4yof Iowa received transdermalfmtunyl afler discontinuingtheirprior optid anakesic; ? p&?&s completed questionnaires before and at 2 and 4 wkfolhxuing transdermalf~tanyl application. ?here was no s&n$&nt d@rence in vimaL analogue scale scoresforpain or mood. Erbalpain descr$tor scores improved at 2 wk (P c .05). lltere was a nons&$cant tenden~ toward increased d$ression and nauseq’ however, patients spent less t&e thinking about their ii&ss andfet their cancer was Lessdt&uptiueto their closest j%uWrelatiues. Constipation, appetite, drowsiness, and concentration were not stahSal& d@rent. Pati reported improved sleep habits at 2 wk (P < .05) and tended to require less help z&h eating, drRtrbrg,washing, and using the bathroom. Allpa&& completi~ the sttiy chose to continue transdermal fentanylfm their cancer pain management. In summcuy, these data demonstrate the anaigesic efiq of the transdermalfentanyl system and suggest that some patients with cancer-relaied pain could benejtjom its ue. J Pain Symptom Manage 1992;7:S58+62.
&YWd
Pain, titeraH, cancer, narcotics
Introduction Appropriate management of cancer-related pain requires 1) adequate knowledge of the pain mechanism(s) involved, 2) inclusive pain assessment,’ 3) identification of barriers to adequate pain control,* and 4) a multifaceted approach to pain therapy. For some patients, current therapeutic options cause signilicant adverse effects, are inconvenient, provide inadequate analgesia, or are expensive. Transdermal fentanyl is a novel addition to the armamentarium against cancer-related pain. Small surveys suggest that some patients treated with Ada& reprintrequeststo: Timothy J. Maves, MD, Pain Clinic, Department of Anesthesia, University of Iowa College of Medicine, Iowa City, IA 52242. @ U.S. Cancer Pain Relief Committee, 1992 Published by Elsevier, New York, New York
transdermal feutanyl will experience relatively fewer side effects3 and have steady-state fentanyl blood levels that fluctuate less than with opioids delivered by many other routes.4 The transdermal route also may be convenient for patients and cost competitive with other noninvasive therapeutic options. As part of a multicenter study, we examined the effect of transdermal fentanyl on the quality of lie of cancer patients.
Methods The experimental protocol was reviewed and approved by the University of Iowa Hospitals and Clinics Human Subjects Review Committee. A total of 10 cancer patients received transdermal fentanyl. Inclusion criteria included a histologically confirmed malignancy, pain related to the cancer
om-3924/92/85.00
til. 7.h.
3 (Sub@) April 1992
I’huse IV Trial ojTransderma1 Fentanyl
and/or therapy, pain necessitating the use ofopioid analgesics, age of 18 yr or older, and ability to communicate effectively with study personnel. Following informed consent, each patient completed 1) visual analogue scale (MS) for pain intensity and mood, 2) verbal pain descriptor (VPD, series of words describing the intensity of pain, with each word assigned a numerical score), 3) Functional Living Index--Cancer (FLJC), and 4) a symptom questionnaire (SQ). History, physical examination, and appropriate laboratory studies were obtained for each patient. Patients previously requiring opioid analgesics were converted to an intramuscular (IMJ morphine equivalent dose using the equianalgesic potency conversions reported by F01ey.~ The analgesic activity ratio of 10 mg IM morphine every 4 hr to 100 pg intravenous (IV) fentanyl every hour was used to derive the initial transdermal fentanyl dose. Prior opioid analgesics were discontinued, and all sedative-agent do.sages were halved when transdennal fentanyl was applied. Opioid-naive patients .were started on 25 gg/hr transdermal fentanyl. Breakthrough pain was treated with short-acting oral opioids (10 mg
s59
morphine or %mg hy~omo~hone) every 20 min as needed for pain control. TransdermaS fentanyl dosage was increased every 2-3 days as indicated by the daily dosage requirements for breakthrough medication. Patients were followed by phone OF personal interview every 1-3 days. Any side effects were treated promptly with appropriate agents or interventions. VAS, VPD, FLIC, and SQ were completed by each patient 2 and 4 wk following initial transdermal fentanyl application. The questionnaire data were analyzed by calculating the mean f SEM for each question. Data obtained prior to treatment with transdermal fentanyl were compared to data at 2 and 4 wk using paired Student’s t test.
Patient profiles are illustrated in Table I; 7 patients completed the study, 1 patient died 2 wk into the study, a second developed septic shock and transdermal fentanyl was diicontinued, and the third patient was withdrawn for increase; nausea and vomiting during chemotherapy. Tabb I
Patient Profile Initial tKillSdcrmal
Dose at
dose
study
fentanyl end of Cancer
Age (yr)
Sex
type
Prior TX
Prior analgesic
(&hr)
1
45
F
XRT
2
60
M
Lun_g (WDNW Lung
3
51
M
Luns
XRT
4
68
M
‘SC) Lung
XRT
Controlled-release morphine Controlled-rclcasc morph& Controlled-release morphine Controlled-rclcasc
5
70
M
(PDNSC) Sq. Ceil BOT
XRT
Controlled-release
Patient
6
65
F
7
54
F
Breast (Infil. Ductal) Cervical
8
64
M
Prostate
9
52
F
10
48
M
Colon (Adeno) Non-Hodgkin’s lymphoma
aDicd
Chcmo
Chcmo Surgical XRT, Chemo Surgical Chcmo, Rad Surgical XRT$ Hormonal Surgical XRT, Chcmo Surgical Chcmo
at 2 wk elevated liver hnction tests. bDiscontinued at 2 wk titer developing G-sepsis. CDiscontinued at 2 wk secondary to naura.
morphine
(&hr)
Pain types
50
1ooa
25
75
Bone mets, leftbra&al plexopathy Bone mets
75
500b
25
50
25
125
Tumor related
50
100
Left bra&al
50
125
Tumor related
25
150
Bone mets
100
150
Tumor related
Bone mets Bone and liver mets
l7lOl$liIlC
Controlled-release morphine Controlled-release morphine Controlled-release morphine Hydromorphone Controlled-rclcasc morphine
50
5oc
Tumor relp!
plcxopathy
RL. 7No. 3 (Suppl.)April 1992
Mwes and Barcelhs
S60
100 z
E s D .-is P
a0 60 40 20 0
a
I pr9-
2-week
4-week
pre-
P-week
4-week
a
b
b
-
pre-
2-week
4-week
pre-
2-week
4-week
pre-
2-week
4-weetc
Fig. 1. Visual analog scale (mm) for (a) pain (0 = least
possiblepain, 100 = worst possible pain) and (b) mood (0 = best mood, 100 = wont mood) pre-, 2 wk, and 4 wk after transdermal fentanyl application, recorded as mean f SEM.
There was no significant difference in VAS for pain or mood at 2 or 4 wk compared with pretreatment values (Figure 1). Transdermal fentanyl improved analgesia, as evidenced by lower VPD score at 2 wk (PC .05), but VPD score at 4 wk was not statistically different (Figure 2a). Analysis of the FLIC revealed a nonsignificant tendency toward increased depression and nausea (Figure 2b and c, respectively). Although not signifrcandy different, patients spent less time thinking about their illness and felt their cancer was less disruptive to their closest friends/relatives (Figure 3a and b, respectively). There was no significant diirence in constipation, appetite, drowsiness, or concentration when analyzing the SQ Patients did report improved sleep habits at 2 wk (P c .05), but the difference was not significant at 4 wk (Figure 4a). Patients tended to require less help with eating, dressing, washing, and using the bathroom (Figure 4b). All patients completing the study chose to continue transdermal fentanyl for their cancer pain management (Figure 4c).
6 85 $4 !I In 3 3 c 2 1 0 C
Fig. 2. Functional living index for cancer scores: (a) pain description scale (0 = no pain, 7 = excruciating), @) depression scale (0 = never, 7 = continually), and (c) nausea scale (0 = none, 7 = a great deal) pre-, 2 wk, and 4 wk after transdermal fentanyl application, recorded as mean f SEM (*P < .05).
Discussion Steady-state fentanyl blood levels resulting in increased analgesia and uninterrupted sleep could potentially have a great impact on quality of life for cancer patients. Although the data did not reveal marked improvement in functional abilities, many spouses expressed that their wives/husbands were more alert and interactive while using transdermal fentanyl. All patients completing the study expressed a desire to continue transdermal fentanyl for their pain management. Bypassing the gastrointestinal tract during opioid
S61
2-week
$-week
P-week
Q-week
b pre-
pre-
2-week
B-week
pm-
P-week
$-week
Fig. 3. Functionalliving index for cancer scores:(a) time spent thinking ahout their illness (0 = never, 7 = constantly),and (9) cancer-relateddisruptionto relatives (0 = no disruption,7 = totallydisruptive)pre-, 2 wk, and 4 wk after transdermalfentanylapplication,recorded as mean * SEM. administration could contribute to a smaller incidence of nausea and constipation.‘j Iftrue, the small number of patients in our study may account for the lack of improvement in these symptoms. Respiratory depression, pruritus, confusion, and allergic reaction did not occur in our patients during transdermal fentanyl application. Of the 10 patients, 9 required at least a 500/o increase in transdermal fentanyl dose during the study. The dose escalation may have resulted from inadequate analgesia prior to the study, relatively low initial fentanyl dose secondary to a conservative conversion ratio (6:1, PO to IM morphine), progression of disease, and/or tolerance. Opioid tolerance may depend on efficacy of the specific opioid (a high-efficacy agonist may down-regulate fewer receptors and produce less tolerance)7 and method of delivery (it has been suggested that continuous IV or intrathecal routes develop tolerance faster than intermittent oral dosing, but this has not been confiied).8 Tolerance to transdermal fentanyl should be addressed in future studies. In summary,
transder.mal
fentanyl
offers an
G
P-week
4-week
Kg. 4. Symptom questioriaire: (a)frequency at which pain preventeda good night’ssleep(0 = none of the time, 5 = all of the time); (b) frequency at which patient required help eating, dressing,washing, and using the bathroom(0 = none of the time, 5 = all of the time)pre-, 2 wk, and 4 wk tier transdermalfentanyl application, recorded as mean +zSEM CP < .05); (c) percentage of patientswho would use transdermalfentanylagain.
efficacious and convenient alternative to conventional opioid therapy. Data from this open-label trial suggest that some patients could benefit when transdermal fentanyl is utilized in the management of cancer-related pain.
We thank Michael Burcham and Linda Hendricks for preparation of the graphic and Peggy Miller for assistance with patient evaluation.
862
Maves and Bmceh
Refeemes 1. F&y KM. Assessment of pain in patients with cancer. In: Swerdlow M, Ventafiidda V, eds. Cancer pain. Norwell, MA: MTP Press, 1987~37-57. 2. Ieti DN, Cleeland CS, Dar R Public attitudes toward cancer pain. Cancer 1985;56:2337-2339. 3. Simmons MA, Payne R, Richenbacher J, Moran K, Southam MA. ‘ITS (fentanyl) in the management of pain in patients with cancer. Proc Am Sot Clin Oncology 1989;8:234. 4. Miser AW, Narang PK, Dothage JA, Young RC, Smdelar W, Miser JS. Transdermal fcntanyl for pain
Vol. 7flo. 3 (Supp f.) Afnif 1992
control in patients with cancer. Pain 1989;37: 15-2 1. 5. Foley KM. The treatment of cancer pain. N Engl J Med 1985;313:84-95. 6. Ventaftidda V, Spoldi E, Caraceni A, Tamburini M, DeConno F. The importance of continuous subcutaneous morphine admiistration for cancer pain control. Pain Chn 1986; 1:47-55. 7. Sosnowski M, Yaksh TL. Differential cross-tolerance between intrathecal morphine and sufentanil in the rat. Anesthesiology 1990;73:1141-1147. 8. JaffeJH. Drug addiction and drug abuse. In: Goodman GA, Goodman Is, Rail TW, Murad F, eds. The pharmacological basis of therapeutics. New York Pergamon Press, 1990:533.