- Email: [email protected]
Safety and efficacy of continuous intravenous morphine for severe cancer pain
CLINICAL STUDIES
Safety and Efficacy of Continuous Intravenous Morphine for Severe Cancer Pain
MARC L. CITRON, M.D.’ ANITA JOHNSTON-EARLY, R.N. BYRON E. FOSSIECK, Jr., M.D.7 STEVEN H. KRASNOW, M.D. ROBIN FRANKLIN, R.Ph. SAMUEL V. SPAGNOLO, M.D. MARTIN H. COHEN, M.D. Washlngon, 0. C.
TO study the efficacy and safety of continuousiy administered intravenous morphine for carker pain uneiieved by standard narcotic therapy, boius intravenous InJections of 2 to 5 mg of morphine were given every 10 minutes until pain relief was achkved. Wiih+nthe next hour, continuous intravenous morphine infusion was begun with the hourly dose equal to the cumulative boius dose. Respiratory rate, puke, blood pressure, arterial biood gas vaiues, mental status, and pain relief were recorded at baseline and during the study period. A reduction in arterial oxygen pressure (Paol) and/or increase in arterial carbon dioxide pressure Pacoz of more than 20 percent of baseline values occurred, during the first 24 hours of infusion, in a minority of patients. This did not requke changes in hourly morphine dose. Despite subsequent increases in morphine dose, blood gas values tended to remain at or return toward basetine values. Severe toxicity occurred during one trial and was heralded by bradypnea and marked somnolence. MaJor pain relief was achieved in 11 of 15 trials. Therefore, continuous intravenous morphine is effective and safe therapy. Bradypnea associated with marked somnolence is a cause for dose reduction. More than 60 percent of patients with cancer will, at some time have pain that requires active medical, surgical, or other intervention [ 11. initially, oral narcotics may provide effective analgesia for moderate to severe pain. As disease progresses, however, oral medications may become ineffective or poorly tolerated. In such patients, pain may also be refractory to parenteral narcotics or may require frequent intramuscular injections. Continuous intravenous morphine has been reported to be beneficial in these circumstances [2-71. To date, however, there has been no evaluation of potential cardiopulmonary or central nervous system toxicities of this mode of therapy. This paper reports the prospective evaluation of the efficacy and safety of continuous intravenous morphine in patients with advanced cancer who had intractable pain.
From the Sections of MedIcal Oncology and Pul-
monaryDiseases, VeteransAdministr%ionfvtedical Center, WashIngton,DC. This work was supportedin part bythe ResearchService of the Veterans Admlnlstratlan.Manuscriptaccepted January24,1994. lCtxrentad&essanded&est3forreprfntrquests: Divislon of Hemstokqy-Dncolosy,Long Island Jewlsh-tiiltskls MedIcalCenter,New Hyds Park, New York 11042. t Died on June 9, 1993.
PATiENlS AND METHODS Patients elfgible for study hada histologically proved diagnosis of cancer and severe pain not relieved by conventionalmethods of narcotic administration. Informed consent, approved by the Medical Center Human Studies Subcommittee, was obtained from all patients. Respiratory rate, pulse, blood pressure, arterial blood gas values, pain intensity, pain relief, and mental status were recordedat baseline and during the study period. Bolus intravenous injections of 2 or 5 mg of morphine were given every 10 minutes, into the side-arm of a running intravenous infusion until patients reported relief. The 5 mg bolus dose was selected for patients
August 1984
The AmericanJournal of Medtclne Volume77
199
CONTINUOUS
TABLE I
MORPtllNE INFUSION4ITRON
ET AL
Baseline Arterial Blood Gas Values, Respiratory Rate, Systolic Blood Pressure, and Pulse Rate before Continuous Intravenous Morphine Infusion in 15 Trials Respiratory
Trial NUllher
Rate
Pao2
PaO0,
(W
Wr)
PH
Group I
Wr)
Pulse (perminute)
1
65
2 3 4 5
65 63 63 60
42 33 43 32 48
7.42 7.40 7.43 7.52 7.39
22 20 24 18 14
108 132 122 110 150
90 100 100 104 100
84 78 94 93 75 77 74 84 101 83
43 45 33 41 36 41 38 34 41 44
7.41 7.41 7.45 7.38 7.37 7.43 7.47 7.46 7.41 7.36
20 16 20 18 13 28 28 18 24 12
156 130 174 110 110 140 120 136 138 140
88 96 96 96 140 100 100 84 108 92
Group11 6 7 8 9 10 11 12 13 14 15
in severe distress. Vitalsigns were determined prior to each
intravenous bolus dose of morphine. Within the next hour, a morphine infusion via an IVAC pump was begun, with the hourly dose equal to the cumulative bolus dose. After morphine infusion was begun, vital signs were determined every 15 minutes for one hour and then every two hws. After the first day of infusion, the hourly morphine dose was increased up to 33 percent daily if the patient had inadequate pain relief and absence of significant toxicity. Morphine solutions were prepared according to a previously described technique [7]. The patient was asked to report intensity and relief of pain, respectively, by the following scale: none, 0; slight, 7; moderate, 2; and severe (for pain) or complete (for relief), 3. The patient’s sedation was evaluated by the observer on the following scale: none (patient alert), 0; mild (patient drowsy on occasion but easily roused), 1; moderate (patient often drowsy but easily roused), 2; severe (patient somnolent and difficult to arouse), 3. Arterial blood gas values were determined at baseline, four hours after onset of infusion, and then daily over the next three days. Patients receiving a morphine infusion for longer than four days then had arterial blood gas determinations weekly or more often if clinically indicated. RESULTS
Thirteeh consecutive patients with a mean age of 52 years (range 32 to 60) underwent 15 courses of continuous intravenous morphine. All patients had normal serum creatinine levels; four trials were undertaken in patients with liver dysfunction, which was defined as elevations in serum aminotransferases and alkaline pfvJspha@e. Pretreatment arterial blood gas values and vital signs are shown in Table I. In five trials, the baseline arterial oxygen pressure (Paon) values were 65 torr
200
(perminute)
Blood
Pressure
August 1884
The American Journal of MedlcJne
or less (Group I). The other 10, with a higher Pao,, were designated Group II. In two trials, the prestudy arterial carbon dioxide pressure (Pacon) was 45 torr or more. Mean systolic blood pressures and pulse rates for all patients were 131 torr (range 108 to 174) and 100 per minute (range 84 to 740), respective/y. Changes in Pao, and Pacon and hourly morphine dose for the five trials in Group I for the first 96 hours of infusion are shown in Figure 1. During four trials, Paon values were within 10 percent of baseline after 24 hours of infusion, whereas during the fifth (Trial 3), the Pao, decreased by 20 percent (absolute fall 13 torr). After 96 hours, the Paoz was still within baseline for three of the five trials. Regarding the other two trials, oxygen administration was required in one (Trial 5), and the Paon was 9 torr below baseline in another (Trial 3). As shown in Figure 1, the Pace, values remained within 11 percent of baseline during the first 24 hours in three of the five trials. However, during the two other (Trials 3 and 4), the Pacop increased 21 percent above baseline (absolute values 5 and 7 torr respectively). The patients’ mental status was unchanged and no remedial action was undertaken. By 96 hours, all P%on values were within 12 percent of baseline. In Trial 5, baseline Pacon was 48 torr and the 96-hour value was 47 torr. The mean hourly morphine doses for these five trials at four, 24, and 96 hours were 20 mg per hour (range 10 to 30 mg), 20 mg per hour (range 6 to 34 mg), and 33 mg per hour (range 20 to 60 mg), respectively. Changes in Pao2 and Pko, and hourly morphine dose for the 10 trials in Group II for the first 96 hours of infusion are shown in Figure 2. Arterial blood gas values
Volume 77
COPJWWOUS
h4ORpHM
HFUSION--C~~‘RON
ET AL
100 0 =Mean
80 j
Pao2
60
(Ton), ’ PaC02 4. 1 -r norphine _I. ng/hour
20
-1 20 -8 ‘0 -4 0
Fmre 1. &oup i: Pa&, Paca, and morphhe doses during first four days of ._ infusion.
Y, 04
--__
r..-.
-
_-_
I 48
t 24
HOURS ____..
-._
1 72
’ A0 96
_.---.
~..-.-_..~.-~
---
--
.-. - __d
.-___
120 100
Pa02
80
VW
60
PaCC2
4~
Flgun 2. Group II: Paot, Pace. and morphine doses during first four &ys of infusion.
were determined at fou hours in eight trials. During five trials, the Pao, values were within 5 percent of baseline at four hours of infusion. During three (Trials 6, 8, and IO), the Pao* decreased by more than 20 percent at four hours. The first of these patients showed marked somnolence and a 70 percent reduction in respiratory rate, which necessitated termination of the morphine infusion, but did not require treatment with a narcotic antagonist. The second patient had no change in his
24
04
.___-
72
-.
HOYRS ._--.. _
96
-.-
_--.
._
J
mental status or respiratory rate, and hi Pao, increased by 14 percent over the next 20 hours witfr no alteration in the morphine dose. In the third, sepsis developed, and the patient died within 24 hours of infusion. In the two trials in which arterial blood .gas values were not determined at four hours, oxygen values were within 15 percent of baseline at 24 and 96 hours. As shown in Figure 2, the Pea, values remained within 12 percent of baseline during the first 24 hours
1984
T?mAmerlcsnJowndd~
volufna
77
201
CONlWWWS
TABLE II
MORPHINE INFUSION--CITRON ET AL
Total Study Days (TSD), Days to Peak Hourly Doss (DTP), Peak Hourty Doss (PD) Arterlal Blood Gas Values, and Resplratofy Rate at Peak Hourly Morphine Dose for Six Contlnuous Intravenous Morphihe Trials Lastlng at Least 14 Days
Trial TSD
DTP
3
80
5 8 9 11 14
32 22 26 25 55
10 17 16 2 11 15
;zj
2% 140 100 20 36 70 75
65 58 70 77 70 ND
pacoa (W 47 40 40 47 45 ND
RWatw Rate PH 7.41 7.34 7.40 7.35 7.44 ND
(per mhW 16 16 9 8 24 18
ND = not determined at peak dose.
in eight of the 10 trials. However, during the other two trials previously mentioned (Trials 6 and 8), the Pacon increased to 20 percent above baseline, 8 and 7 torr respectively. The morphine infusion was discontinued in Trial 6 and continued in Trial 8. By 96 hours, the Pace, was within 5 percent of baseline values in all but one case in which it was elevated 12 percent over baseline (Trial 14). In Trial 7, the baseline Pacoz was 45 torr, and the 96-hour value was 44 torr. Dose decrements were required in only one trial during the first 96 hours; in seven others, the hourly morphine dose was increased above that given at 24 hours. The mean hourly morphine doses for these 10 trials at four, 24, and 96 hours were 20 mg (range 4 to 50 mg), 31 mg (range 4 to 66 mg), and 50 mg (range 8 to 130 mg), respectively. Among all trials, there was no direct correlation between respiratory rate and arterial blood gas values. In seven trials, reductions in respiratory fate were associated with a decrease in Pao,, and in three trials, there was an increased respiratory rate with a decrease in Paon. In three trials, there was an increase in Pao, associated with a decrease in respiratory rate, and in two, there was a Pao, decrease without a change in respiratory rate. However, in six of seven patients with a baseline respiratory rate of 20 or less, a reduction in Pao, occurred when breathing frequency dropped more than 40 percent from baseline. Data for six patients who received continous intravenous morphine for longer than two weeks are shown in Table II. Arterial oxygen values remained within a mean of 12 percent (range 3 to 25 percent) of pretreatment values, whereas Pace, values increased to a mean of onty 10 percent (range 0 to 2 1 percent) above baseline values. These patients received a mean peak dose of 61 mg morphine per hour, which was 400 percent greater than their mean dose at 24 hours. The highest dose of morphine given was 359 mg per hour
202
August 1984
The American Journal ot hbdlclne
in Trial 15 to a man with advanced small cell carcinoma. His infusion lasted eight days, and the peak dose was reached on the seventh day. Arterial blood gas determinations when the morphine dose was 210 mg per hour revealed a Paon of 71 ton and Pace, of 55 torr with the patient receiving 1 liter of nasal oxygen. He died on the eighth day due to his metastatic cancer. Systolic blood pressures fell a mean of 15 percent from baseline in 11 trials at 96 hours, but clinically significant morphine-induced hypotension did not occur in any trial. No consistent changes in pulse rate were associated with morphine treatment. Within the first 96 hours, pulse rates remained within 20 percent of baseline for all surviving patients, and no episodes of morphine-induced bradycardia or tachyarrythmia were encountered. During the first 96 hours in nine trials, patients were scored as 0 (alert) or 1 (slightly drowsy) in the sedation rating scale. Two patients, however, showed unexpected severe somnolence on treatment. During Trial 4, the patient become somnolent and had garbled speech when aroused. Arterial blood gas values were stable and laboratory results revealed hypercalcemia. With treatment of the hypercalcemia, he became more alert. During Trial 6, as already mentioned, the patient exhibited a marked decrease in Pao,, decrease in respiratory rate, and increasing somnolence starting 12 hours after initiation of morphine. Initial morphine doses were titrated to relieve pain at rest. Once this was accomplished, further increments were made to allow for pain relief with ambulation. Moderate to complete pain relief was attained during 11 trials: slight to moderate pain relief was produced in the four other trials. These latter patients had predominant neuritic pain components in addition to multiple sites of painful metastatic disease. Despite the subsequent use of nerve blocks and neurosurgical procedures, these patients, although comfortable, did
Volume 77
not experience sustained major pain relief. Peak hourly morphine doses given during the latter trials were 36 mg, 140 mg, 200 mg, and 310 mg per hour. Of 13 patients treated in this study, four deathsthree due to cancer and one due to sepsis-occurred within the first week of therapy. No patient died as direct result of the administration of morphine. Ten other patients were treated for a total of more than 300 study days. Two patients (Trials 8 and 9) included in the stwfy because of inability to take oral medications, underwent gastrostomy and then were able to receive enteral analgesics. Three patients (Trials 3,14, and 15) underwent nerve blocks for pain control but continued to require morphine infusion. Three patients (Trials 3, 5, and 8) received palliative radiotherapy fcf pain while receiving continuous morphine. One patient underwent rhizotomy for pain control. Five patients after six trials (Trials 1, 2,6 to 9, and 11) were discharged from the hospital with oral or enteral analgesics, COMMENTS These data indicate that continuous intravenous morphine is a safe and effective means of relieving pain, even in persons with borderline pulmonary status (Pao, of 65 or less and Pacop of 45 or more). Although continuous morphine generally caused a reduction in Pao, in the first 24 hours of treatment, this was generally not associated with an increase in Pace, or a decline in mental status. As the infusion was continued for 96 hours and beyond, arterial blood gas values tended to remain sufficiently near baseline so as not to require clinical intervention. Similarly, after the initiation of the morphine infusion, pulse rates, and systolic blood pressures did not change significantly from their baseline values. Respiratory rate tended to decrease during continuous morphine infusion. There was not, however, a uniform association between changes in arterial blood gas values and respiratory rate for the first 96 hours. For example, patients with a baseline respiratory rate of 20 or less who had more than a 40 percent rate decrease from baseline had resultant Pao, reductions. In contrast, for patients with respiratory rates greater than 20, there was no clear correlation with changes in Pao,. This may have been due to other supervening conditions such as heart failure, pneumonia, or metastatic disease to the lung that occurred in these patients. It is not clear, moreover, why continuous morphine infusion produced more frequent and greater Pao, reductions than Pace, elevations. This phenomenon may be explained, in part, by an effect on ventilation-perfusion factors in addition to morphine’s known actions on the respiratory Center [ 81, chemoreceptor threshold for Pacon [ 91, and pulmonary J-receptors [ lo].
The occurrence of bradypnea without a deterioration in mental status was well tolerated. Morphine-induced life-threatening toxicity, however, was characterized clinically by bradypnea in the presence of declining alertness. Prior to development of grade 3 sedation (somnolent and difficult to arouse), the patient may have slow and garbled speech. Bradypnea associated with marked somnolence was a cause for dose reduction. Likewise, the onset of somnolence without bradypnea should raise the possibility of other causative factors such as hypercalcemia. On the basis of these results, we believe that continuous intravenous morphine can be adequately monitored by following respiratory rate (in persons with pretreatment rates of 20 per minute or less) and mental status. In such persons, a 40 percent decrease or more in respiratory rate attended either by clinical manifestations of hypoxia or by somnolence indicates the need for arterial blood gas determinations. Morphine doses in this study were initially adjusted to produce pain relief at bedrest. The hourly rate was then escalated as the patients’ activity level increased. Despite further increases in morphine doses, patients were able to walk, eat, and talk without undue sedation. However, development of tolerance to the drug’s analgesic action may also have been a factor necessitating dose increments. The common use of dose escalation in this study contrasts with one report in which this was not needed to control pain [ 51. In some of our trials, it took more than one week to arrive at an optimal morphine dose. The majority of patients treated experienced major to complete relief during their course. Dther patients who experienced slight to moderate relief were given palliative radiotherapy, nerve blocks, and neurosurgical procedures. In patients with marked neuritic components due to nerve plexus involvement, these efforts to control pain were of limited benefit and did not allow major reductions in morphine dose. When oral analgesics are not effective or feasible, parenteral narcotics are required to relieve pain. Unfortunately, severe cancer pain is frequently treated inadequately, and patients may experience extreme distress while waiting for injections. To determine whether continuous morphine is superior to intermittent narcotic injections would require a prospective study of those two treatment methods. This study, however, supports a growing body of literature indicating that continuous intravenous morphine is safe and effective treatment for severe cancer pain. ACKNOWLEDGMENT We wish to acknowledge the assistance of Mrs. Rosemary Giles in the preparation of this manuscript.
Aqusll994
The Amerkan Journal of Medkine
Volume 77
203
CONTINUOUS
htORPHM
lNFUSlON-CllRON
ET AL
REFERENCES 1.
2. 3. 4. 5. 6.
204
Bonica JJ: Importance of the problem. In: Bonica JJ, Ventafrkfda V, eds. Advances in pain research and therapy, vol 2. New York: Raven Press, 1979; 1-12. Holmes AH: Morphine IV infusion for chronic pain. Drug Intel1 Clin Pftarm 1978; 12: 558-557. Ensworth S: Morphine IV infusion for chronic pain (letter). Drug lntell Clin f%arm 1979; 13: 297. Kowolenko M: An addftionalccmment on rraxphine IV infusion. Drug lntell Clin Pharm 1980; 14: 296-297. Miser AW, Miser JS, Clark BS: Continuous IV infusion of morphine sulfate fa control of severe pain in children with terminal malignancy. J Pediatr 1980; 98: 930-932. Fraser DC: Intravenous morphine infusion for chronic pain. Ann Intern Med 1980; 93: 781-782.
Augusll994
The Amarkan Journal of Mdklne
Volume 77
7.
8.
9.
10.
DeChristoforo R, C&en BJ, Hood JC, Narang PK, Magratft IT: High-dose morphine infusion complicated by chlorarnbutol-Induced somnolence. Ann Intern hkd 1983; 98: 335-336. Florez J, Borrison HL: Effects of central depressant drugs on respiratory regulation in the decerebrate cat. RespfrPhysid 1969; 6: 318-329. Ngai SH: Effects of morphine and meperidine on the central respiratory mechanisms in the cat: the action of levallorphan in antagonizing these effects. J Pftarmacol Exp Ther 1961; 131: 91-102. Willette RN, Sapru HN: Peripheral versus central cardiorespiratory effects of morphine. Neurophamtacology 1982; 21: 1019-1026.
Safety and Efficacy of Continuous Intravenous Morphine for Severe Cancer Pain
MARC L. CITRON, M.D.’ ANITA JOHNSTON-EARLY, R.N. BYRON E. FOSSIECK, Jr., M.D.7 STEVEN H. KRASNOW, M.D. ROBIN FRANKLIN, R.Ph. SAMUEL V. SPAGNOLO, M.D. MARTIN H. COHEN, M.D. Washlngon, 0. C.
TO study the efficacy and safety of continuousiy administered intravenous morphine for carker pain uneiieved by standard narcotic therapy, boius intravenous InJections of 2 to 5 mg of morphine were given every 10 minutes until pain relief was achkved. Wiih+nthe next hour, continuous intravenous morphine infusion was begun with the hourly dose equal to the cumulative boius dose. Respiratory rate, puke, blood pressure, arterial biood gas vaiues, mental status, and pain relief were recorded at baseline and during the study period. A reduction in arterial oxygen pressure (Paol) and/or increase in arterial carbon dioxide pressure Pacoz of more than 20 percent of baseline values occurred, during the first 24 hours of infusion, in a minority of patients. This did not requke changes in hourly morphine dose. Despite subsequent increases in morphine dose, blood gas values tended to remain at or return toward basetine values. Severe toxicity occurred during one trial and was heralded by bradypnea and marked somnolence. MaJor pain relief was achieved in 11 of 15 trials. Therefore, continuous intravenous morphine is effective and safe therapy. Bradypnea associated with marked somnolence is a cause for dose reduction. More than 60 percent of patients with cancer will, at some time have pain that requires active medical, surgical, or other intervention [ 11. initially, oral narcotics may provide effective analgesia for moderate to severe pain. As disease progresses, however, oral medications may become ineffective or poorly tolerated. In such patients, pain may also be refractory to parenteral narcotics or may require frequent intramuscular injections. Continuous intravenous morphine has been reported to be beneficial in these circumstances [2-71. To date, however, there has been no evaluation of potential cardiopulmonary or central nervous system toxicities of this mode of therapy. This paper reports the prospective evaluation of the efficacy and safety of continuous intravenous morphine in patients with advanced cancer who had intractable pain.
From the Sections of MedIcal Oncology and Pul-
monaryDiseases, VeteransAdministr%ionfvtedical Center, WashIngton,DC. This work was supportedin part bythe ResearchService of the Veterans Admlnlstratlan.Manuscriptaccepted January24,1994. lCtxrentad&essanded&est3forreprfntrquests: Divislon of Hemstokqy-Dncolosy,Long Island Jewlsh-tiiltskls MedIcalCenter,New Hyds Park, New York 11042. t Died on June 9, 1993.
PATiENlS AND METHODS Patients elfgible for study hada histologically proved diagnosis of cancer and severe pain not relieved by conventionalmethods of narcotic administration. Informed consent, approved by the Medical Center Human Studies Subcommittee, was obtained from all patients. Respiratory rate, pulse, blood pressure, arterial blood gas values, pain intensity, pain relief, and mental status were recordedat baseline and during the study period. Bolus intravenous injections of 2 or 5 mg of morphine were given every 10 minutes, into the side-arm of a running intravenous infusion until patients reported relief. The 5 mg bolus dose was selected for patients
August 1984
The AmericanJournal of Medtclne Volume77
199
CONTINUOUS
TABLE I
MORPtllNE INFUSION4ITRON
ET AL
Baseline Arterial Blood Gas Values, Respiratory Rate, Systolic Blood Pressure, and Pulse Rate before Continuous Intravenous Morphine Infusion in 15 Trials Respiratory
Trial NUllher
Rate
Pao2
PaO0,
(W
Wr)
PH
Group I
Wr)
Pulse (perminute)
1
65
2 3 4 5
65 63 63 60
42 33 43 32 48
7.42 7.40 7.43 7.52 7.39
22 20 24 18 14
108 132 122 110 150
90 100 100 104 100
84 78 94 93 75 77 74 84 101 83
43 45 33 41 36 41 38 34 41 44
7.41 7.41 7.45 7.38 7.37 7.43 7.47 7.46 7.41 7.36
20 16 20 18 13 28 28 18 24 12
156 130 174 110 110 140 120 136 138 140
88 96 96 96 140 100 100 84 108 92
Group11 6 7 8 9 10 11 12 13 14 15
in severe distress. Vitalsigns were determined prior to each
intravenous bolus dose of morphine. Within the next hour, a morphine infusion via an IVAC pump was begun, with the hourly dose equal to the cumulative bolus dose. After morphine infusion was begun, vital signs were determined every 15 minutes for one hour and then every two hws. After the first day of infusion, the hourly morphine dose was increased up to 33 percent daily if the patient had inadequate pain relief and absence of significant toxicity. Morphine solutions were prepared according to a previously described technique [7]. The patient was asked to report intensity and relief of pain, respectively, by the following scale: none, 0; slight, 7; moderate, 2; and severe (for pain) or complete (for relief), 3. The patient’s sedation was evaluated by the observer on the following scale: none (patient alert), 0; mild (patient drowsy on occasion but easily roused), 1; moderate (patient often drowsy but easily roused), 2; severe (patient somnolent and difficult to arouse), 3. Arterial blood gas values were determined at baseline, four hours after onset of infusion, and then daily over the next three days. Patients receiving a morphine infusion for longer than four days then had arterial blood gas determinations weekly or more often if clinically indicated. RESULTS
Thirteeh consecutive patients with a mean age of 52 years (range 32 to 60) underwent 15 courses of continuous intravenous morphine. All patients had normal serum creatinine levels; four trials were undertaken in patients with liver dysfunction, which was defined as elevations in serum aminotransferases and alkaline pfvJspha@e. Pretreatment arterial blood gas values and vital signs are shown in Table I. In five trials, the baseline arterial oxygen pressure (Paon) values were 65 torr
200
(perminute)
Blood
Pressure
August 1884
The American Journal of MedlcJne
or less (Group I). The other 10, with a higher Pao,, were designated Group II. In two trials, the prestudy arterial carbon dioxide pressure (Pacon) was 45 torr or more. Mean systolic blood pressures and pulse rates for all patients were 131 torr (range 108 to 174) and 100 per minute (range 84 to 740), respective/y. Changes in Pao, and Pacon and hourly morphine dose for the five trials in Group I for the first 96 hours of infusion are shown in Figure 1. During four trials, Paon values were within 10 percent of baseline after 24 hours of infusion, whereas during the fifth (Trial 3), the Pao, decreased by 20 percent (absolute fall 13 torr). After 96 hours, the Paoz was still within baseline for three of the five trials. Regarding the other two trials, oxygen administration was required in one (Trial 5), and the Paon was 9 torr below baseline in another (Trial 3). As shown in Figure 1, the Pace, values remained within 11 percent of baseline during the first 24 hours in three of the five trials. However, during the two other (Trials 3 and 4), the Pacop increased 21 percent above baseline (absolute values 5 and 7 torr respectively). The patients’ mental status was unchanged and no remedial action was undertaken. By 96 hours, all P%on values were within 12 percent of baseline. In Trial 5, baseline Pacon was 48 torr and the 96-hour value was 47 torr. The mean hourly morphine doses for these five trials at four, 24, and 96 hours were 20 mg per hour (range 10 to 30 mg), 20 mg per hour (range 6 to 34 mg), and 33 mg per hour (range 20 to 60 mg), respectively. Changes in Pao2 and Pko, and hourly morphine dose for the 10 trials in Group II for the first 96 hours of infusion are shown in Figure 2. Arterial blood gas values
Volume 77
COPJWWOUS
h4ORpHM
HFUSION--C~~‘RON
ET AL
100 0 =Mean
80 j
Pao2
60
(Ton), ’ PaC02 4. 1 -r norphine _I. ng/hour
20
-1 20 -8 ‘0 -4 0
Fmre 1. &oup i: Pa&, Paca, and morphhe doses during first four days of ._ infusion.
Y, 04
--__
r..-.
-
_-_
I 48
t 24
HOURS ____..
-._
1 72
’ A0 96
_.---.
~..-.-_..~.-~
---
--
.-. - __d
.-___
120 100
Pa02
80
VW
60
PaCC2
4~
Flgun 2. Group II: Paot, Pace. and morphine doses during first four &ys of infusion.
were determined at fou hours in eight trials. During five trials, the Pao, values were within 5 percent of baseline at four hours of infusion. During three (Trials 6, 8, and IO), the Pao* decreased by more than 20 percent at four hours. The first of these patients showed marked somnolence and a 70 percent reduction in respiratory rate, which necessitated termination of the morphine infusion, but did not require treatment with a narcotic antagonist. The second patient had no change in his
24
04
.___-
72
-.
HOYRS ._--.. _
96
-.-
_--.
._
J
mental status or respiratory rate, and hi Pao, increased by 14 percent over the next 20 hours witfr no alteration in the morphine dose. In the third, sepsis developed, and the patient died within 24 hours of infusion. In the two trials in which arterial blood .gas values were not determined at four hours, oxygen values were within 15 percent of baseline at 24 and 96 hours. As shown in Figure 2, the Pea, values remained within 12 percent of baseline during the first 24 hours
1984
T?mAmerlcsnJowndd~
volufna
77
201
CONlWWWS
TABLE II
MORPHINE INFUSION--CITRON ET AL
Total Study Days (TSD), Days to Peak Hourly Doss (DTP), Peak Hourty Doss (PD) Arterlal Blood Gas Values, and Resplratofy Rate at Peak Hourly Morphine Dose for Six Contlnuous Intravenous Morphihe Trials Lastlng at Least 14 Days
Trial TSD
DTP
3
80
5 8 9 11 14
32 22 26 25 55
10 17 16 2 11 15
;zj
2% 140 100 20 36 70 75
65 58 70 77 70 ND
pacoa (W 47 40 40 47 45 ND
RWatw Rate PH 7.41 7.34 7.40 7.35 7.44 ND
(per mhW 16 16 9 8 24 18
ND = not determined at peak dose.
in eight of the 10 trials. However, during the other two trials previously mentioned (Trials 6 and 8), the Pacon increased to 20 percent above baseline, 8 and 7 torr respectively. The morphine infusion was discontinued in Trial 6 and continued in Trial 8. By 96 hours, the Pace, was within 5 percent of baseline values in all but one case in which it was elevated 12 percent over baseline (Trial 14). In Trial 7, the baseline Pacoz was 45 torr, and the 96-hour value was 44 torr. Dose decrements were required in only one trial during the first 96 hours; in seven others, the hourly morphine dose was increased above that given at 24 hours. The mean hourly morphine doses for these 10 trials at four, 24, and 96 hours were 20 mg (range 4 to 50 mg), 31 mg (range 4 to 66 mg), and 50 mg (range 8 to 130 mg), respectively. Among all trials, there was no direct correlation between respiratory rate and arterial blood gas values. In seven trials, reductions in respiratory fate were associated with a decrease in Pao,, and in three trials, there was an increased respiratory rate with a decrease in Paon. In three trials, there was an increase in Pao, associated with a decrease in respiratory rate, and in two, there was a Pao, decrease without a change in respiratory rate. However, in six of seven patients with a baseline respiratory rate of 20 or less, a reduction in Pao, occurred when breathing frequency dropped more than 40 percent from baseline. Data for six patients who received continous intravenous morphine for longer than two weeks are shown in Table II. Arterial oxygen values remained within a mean of 12 percent (range 3 to 25 percent) of pretreatment values, whereas Pace, values increased to a mean of onty 10 percent (range 0 to 2 1 percent) above baseline values. These patients received a mean peak dose of 61 mg morphine per hour, which was 400 percent greater than their mean dose at 24 hours. The highest dose of morphine given was 359 mg per hour
202
August 1984
The American Journal ot hbdlclne
in Trial 15 to a man with advanced small cell carcinoma. His infusion lasted eight days, and the peak dose was reached on the seventh day. Arterial blood gas determinations when the morphine dose was 210 mg per hour revealed a Paon of 71 ton and Pace, of 55 torr with the patient receiving 1 liter of nasal oxygen. He died on the eighth day due to his metastatic cancer. Systolic blood pressures fell a mean of 15 percent from baseline in 11 trials at 96 hours, but clinically significant morphine-induced hypotension did not occur in any trial. No consistent changes in pulse rate were associated with morphine treatment. Within the first 96 hours, pulse rates remained within 20 percent of baseline for all surviving patients, and no episodes of morphine-induced bradycardia or tachyarrythmia were encountered. During the first 96 hours in nine trials, patients were scored as 0 (alert) or 1 (slightly drowsy) in the sedation rating scale. Two patients, however, showed unexpected severe somnolence on treatment. During Trial 4, the patient become somnolent and had garbled speech when aroused. Arterial blood gas values were stable and laboratory results revealed hypercalcemia. With treatment of the hypercalcemia, he became more alert. During Trial 6, as already mentioned, the patient exhibited a marked decrease in Pao,, decrease in respiratory rate, and increasing somnolence starting 12 hours after initiation of morphine. Initial morphine doses were titrated to relieve pain at rest. Once this was accomplished, further increments were made to allow for pain relief with ambulation. Moderate to complete pain relief was attained during 11 trials: slight to moderate pain relief was produced in the four other trials. These latter patients had predominant neuritic pain components in addition to multiple sites of painful metastatic disease. Despite the subsequent use of nerve blocks and neurosurgical procedures, these patients, although comfortable, did
Volume 77
not experience sustained major pain relief. Peak hourly morphine doses given during the latter trials were 36 mg, 140 mg, 200 mg, and 310 mg per hour. Of 13 patients treated in this study, four deathsthree due to cancer and one due to sepsis-occurred within the first week of therapy. No patient died as direct result of the administration of morphine. Ten other patients were treated for a total of more than 300 study days. Two patients (Trials 8 and 9) included in the stwfy because of inability to take oral medications, underwent gastrostomy and then were able to receive enteral analgesics. Three patients (Trials 3,14, and 15) underwent nerve blocks for pain control but continued to require morphine infusion. Three patients (Trials 3, 5, and 8) received palliative radiotherapy fcf pain while receiving continuous morphine. One patient underwent rhizotomy for pain control. Five patients after six trials (Trials 1, 2,6 to 9, and 11) were discharged from the hospital with oral or enteral analgesics, COMMENTS These data indicate that continuous intravenous morphine is a safe and effective means of relieving pain, even in persons with borderline pulmonary status (Pao, of 65 or less and Pacop of 45 or more). Although continuous morphine generally caused a reduction in Pao, in the first 24 hours of treatment, this was generally not associated with an increase in Pace, or a decline in mental status. As the infusion was continued for 96 hours and beyond, arterial blood gas values tended to remain sufficiently near baseline so as not to require clinical intervention. Similarly, after the initiation of the morphine infusion, pulse rates, and systolic blood pressures did not change significantly from their baseline values. Respiratory rate tended to decrease during continuous morphine infusion. There was not, however, a uniform association between changes in arterial blood gas values and respiratory rate for the first 96 hours. For example, patients with a baseline respiratory rate of 20 or less who had more than a 40 percent rate decrease from baseline had resultant Pao, reductions. In contrast, for patients with respiratory rates greater than 20, there was no clear correlation with changes in Pao,. This may have been due to other supervening conditions such as heart failure, pneumonia, or metastatic disease to the lung that occurred in these patients. It is not clear, moreover, why continuous morphine infusion produced more frequent and greater Pao, reductions than Pace, elevations. This phenomenon may be explained, in part, by an effect on ventilation-perfusion factors in addition to morphine’s known actions on the respiratory Center [ 81, chemoreceptor threshold for Pacon [ 91, and pulmonary J-receptors [ lo].
The occurrence of bradypnea without a deterioration in mental status was well tolerated. Morphine-induced life-threatening toxicity, however, was characterized clinically by bradypnea in the presence of declining alertness. Prior to development of grade 3 sedation (somnolent and difficult to arouse), the patient may have slow and garbled speech. Bradypnea associated with marked somnolence was a cause for dose reduction. Likewise, the onset of somnolence without bradypnea should raise the possibility of other causative factors such as hypercalcemia. On the basis of these results, we believe that continuous intravenous morphine can be adequately monitored by following respiratory rate (in persons with pretreatment rates of 20 per minute or less) and mental status. In such persons, a 40 percent decrease or more in respiratory rate attended either by clinical manifestations of hypoxia or by somnolence indicates the need for arterial blood gas determinations. Morphine doses in this study were initially adjusted to produce pain relief at bedrest. The hourly rate was then escalated as the patients’ activity level increased. Despite further increases in morphine doses, patients were able to walk, eat, and talk without undue sedation. However, development of tolerance to the drug’s analgesic action may also have been a factor necessitating dose increments. The common use of dose escalation in this study contrasts with one report in which this was not needed to control pain [ 51. In some of our trials, it took more than one week to arrive at an optimal morphine dose. The majority of patients treated experienced major to complete relief during their course. Dther patients who experienced slight to moderate relief were given palliative radiotherapy, nerve blocks, and neurosurgical procedures. In patients with marked neuritic components due to nerve plexus involvement, these efforts to control pain were of limited benefit and did not allow major reductions in morphine dose. When oral analgesics are not effective or feasible, parenteral narcotics are required to relieve pain. Unfortunately, severe cancer pain is frequently treated inadequately, and patients may experience extreme distress while waiting for injections. To determine whether continuous morphine is superior to intermittent narcotic injections would require a prospective study of those two treatment methods. This study, however, supports a growing body of literature indicating that continuous intravenous morphine is safe and effective treatment for severe cancer pain. ACKNOWLEDGMENT We wish to acknowledge the assistance of Mrs. Rosemary Giles in the preparation of this manuscript.
Aqusll994
The Amerkan Journal of Medkine
Volume 77
203
CONTINUOUS
htORPHM
lNFUSlON-CllRON
ET AL
REFERENCES 1.
2. 3. 4. 5. 6.
204
Bonica JJ: Importance of the problem. In: Bonica JJ, Ventafrkfda V, eds. Advances in pain research and therapy, vol 2. New York: Raven Press, 1979; 1-12. Holmes AH: Morphine IV infusion for chronic pain. Drug Intel1 Clin Pftarm 1978; 12: 558-557. Ensworth S: Morphine IV infusion for chronic pain (letter). Drug lntell Clin f%arm 1979; 13: 297. Kowolenko M: An addftionalccmment on rraxphine IV infusion. Drug lntell Clin Pharm 1980; 14: 296-297. Miser AW, Miser JS, Clark BS: Continuous IV infusion of morphine sulfate fa control of severe pain in children with terminal malignancy. J Pediatr 1980; 98: 930-932. Fraser DC: Intravenous morphine infusion for chronic pain. Ann Intern Med 1980; 93: 781-782.
Augusll994
The Amarkan Journal of Mdklne
Volume 77
7.
8.
9.
10.
DeChristoforo R, C&en BJ, Hood JC, Narang PK, Magratft IT: High-dose morphine infusion complicated by chlorarnbutol-Induced somnolence. Ann Intern hkd 1983; 98: 335-336. Florez J, Borrison HL: Effects of central depressant drugs on respiratory regulation in the decerebrate cat. RespfrPhysid 1969; 6: 318-329. Ngai SH: Effects of morphine and meperidine on the central respiratory mechanisms in the cat: the action of levallorphan in antagonizing these effects. J Pftarmacol Exp Ther 1961; 131: 91-102. Willette RN, Sapru HN: Peripheral versus central cardiorespiratory effects of morphine. Neurophamtacology 1982; 21: 1019-1026.