INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients with chronic obstructive pulmonary disease (COPD) are often limited by disabling breathlessness and may have a poor quality of life. This breathlessness is largely explained by the increased work of breathing that arises from increased respiratory drive and the greater ventilatory load that results from airway obstruction, decreased lung compliance, and dynamic air trapping (1).

Opiates have been advocated as a treatment for breathlessness in COPD, and there are theoretical reasons why they may be beneficial. Opiates reduce ventilatory drive in response to carbon dioxide (4), hypoxia (5), and exercise (6). The decrease in respiratory effort may lead to a reduction in breathlessness. In addition, the sedating effect of opiates may reduce anxiety. There have, however, been few controlled studies of opiates for breathlessness in COPD. In a single-dose study, morphine 0.8 mg/kg given orally 60 min before exercise resulted in an 18% higher maximal work load than placebo, with no increase in breathlessness (7). In another study, a single dose of 1 mg/ kg of dihydrocodeine (DHC) reduced breathlessness at rest by 20% and increased exercise tolerance by 18% over placebo (8). DHC 15 mg, given as much as three times daily before exercise, has also been reported to result in a 16.5% improvement in walking distance and a 12% reduction in breathlessness compared with placebo (9).

In longer-term studies of opiates in COPD the benefits have been less clear. When DHC 30 mg, 60 mg, and placebo were administered three times daily for 2 wk, an improvement in breathlessness over placebo was seen with the 30 mg but not with the 60 mg dose (10). Although a few of the subjects reported considerable benefit, there were significant adverse effects with DHC and five of the 16 subjects did not complete the study because of nausea and vomiting. Two weeks of treatment with diamorphine in doses of 2.5 and 5.0 mg four times daily had no effect on breathlessness or exercise tolerance when compared with placebo (11). These investigators suggested that the doses used may have been subtherapeutic and this could have accounted for the disappointing results. Nebulized morphine has also been investigated as a treatment in COPD. Several single-dose studies have shown no evidence of an effect on breathlessness (12), although one study reported a significant improvement in exercise tolerance (15).

There is uncertainty about the efficacy of morphine for the treatment of breathlessness in patients with severe COPD. To address this question we have studied the effects of morphine sulphate on breathlessness, exercise tolerance, and quality of life in patients with severe COPD. We used sustained-release morphine in preference to morphine elixir because the sustained-release preparation can be used twice a day and this is likely to enhance compliance. We were concerned that the use of too high a dose of morphine initially might lead to unacceptable side effects, so the dose of morphine (or matching placebo) was titrated from 10 mg daily to 20 mg twice a day as tolerated over the first 2 wk of each treatment period. As in other studies on the effects of opiates in COPD, we measured changes in breathlessness and exercise tolerance. In addition, we incorporated a measure of quality of life because we believe that this is an important end point.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Subjects with breathlessness caused by COPD (as defined by the American Thoracic Society) (16) were recruited from outpatient clinics or after they had completed a pulmonary rehabilitation program. Subjects were eligible if their FEV₁ was < 1.5 L and the improvement in FEV₁ after inhaled salbutamol was less than 200 ml. In addition, they were required to have a Pa_O2 > 65 mm Hg (8.5 kPa), a Pa_CO2 < 40 mm Hg (5.5 kPa), and a smoking history of more than 20 pack-years. Subjects were excluded if they had congestive cardiac failure, any significant respiratory disease other than COPD, alcoholism, cognitive impairment, major anxiety disorder, depression, raised liver enzymes (AST > twice the upper limit of normal), serum creatinine > 0.2 mmol/L, inability to perform a 6-min walk, or if they were already receiving opiates. To ensure clinical stability, subjects were not entered in the study if they had required a change in medication in the last month or had been hospitalized within the previous 2 mo. Within the previous year all the subjects had had a chest radiograph consistent with COPD. All subjects gave written informed consent, and the study had the approval of the North Health Ethics Committee.

Study Design

This was a double-blind, placebo-controlled, crossover study, with two 6-wk treatment periods separated by a 2-wk washout. At the initial visit, a history was taken and a physical examination was performed. An arterial blood gas determination was obtained, and venous blood was drawn for a full blood count, electrolytes, and liver function tests. FEV₁ and VC were measured using a dry bellows spirometer (Vitalograph, Buckingham, UK), both before and 20 min after 5 mg nebulized salbutamol. Inhaled bronchodilators were withheld for 4 h prior to the study visit. Subjects also identified five activities important in their everyday life for the Dyspnea subscale of the Chronic Respiratory Disease Questionnaire (CRQ) (17) and undertook a practice 6-min walk (18). The initial visit was followed by a 2-wk run-in period. During the run-in and throughout the study the subjects scored their breathlessness twice a day and recorded this on a diary card. In the evening, the daytime breathlessness score was rated from 0 to 5, with 0 = "no breathlessness at rest or on exertion," 2 = "no breathlessness at rest but breathlessness on mild exertion," and 5 = "severe breathlessness at rest." In the morning, the nighttime breathlessness was scored from 0 to 4, with 0 = "no breathlessness during the night" and 4 = "breathlessness so severe that you did not sleep at all."

Subjects attended the study center at the beginning and end of each 6-wk treatment period. At each of these visits the CRQ was completed and spirometry and a 6MW were performed. During each treatment period the subjects were visited at home on three occasions and were telephoned on four other occasions. On all home visits, as well as on the visits to the study center, we performed pulse oximetry (N-20P; Nellcor Inc., Hayward, CA), enquired about adverse effects, and resupplied medicines and diary cards.

The active treatment consisted of 10-mg tablets of sustained-release morphine sulphate (MST Continus; Douglas Pharmaceuticals, Auckland, NZ) enclosed in size 0 gelatin capsules (Pacific Pharmaceuticals, Auckland, NZ). The placebo capsules were identical in appearance but were filled only with lactose. Randomization was performed by the Auckland Hospital Pharmacy in blocks of four. They maintained the randomization schedule, stored the study medicines in a controlled drug safe, and dispensed them at study visits according to the regulations governing the use of controlled medicines. If there were no adverse effects or they were minor, the dose was titrated in the following fashion: one capsule at night for the first 3 d, increasing to one twice daily after a week, then one in the morning and two at night after a further 3 d, and finally a maximum of two capsules twice daily after 2 wk. There was provision to back titrate if adverse effects became troublesome. Subjects were given the option of continuing on open-label treatment with morphine at the end of the study without breaking the original treatment code.

Statistical Analysis

The primary outcome measure was the CRQ score (17). This instrument consists of a Total Score and four subscales (Dyspnea, Mastery, Fatigue, and Emotional). Secondary outcome measures were distance walked on the standardized 6MW (18), breathlessness scores before and after the 6MW, breathlessness scored twice daily on a diary card, oxygen saturation, spirometry, and side effects. The effect of treatment with morphine on the CRQ, and 6MW distance was assessed using repeated-measures analysis of variance (19). Other parameters were analyzed with paired t tests. When paired t tests were used, the two-tailed p value was reported. When mean scores were given, this was the average with standard error (SE) in parentheses.

The study was designed to have a power of 80% for detecting a 5-point improvement in the CRQ dyspnea scale at the 5% level of significance. On the basis of data from Guyatt and the coworkers (20), we calculated that we would need at least nine subjects in this crossover study.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sixteen subjects (11 males), with a mean age of 70.7 (SE, 1.6) yr and a smoking history of 54 (7.9) pack-years were enrolled in the trial. Twelve were recruited from outpatient clinics and four from an outpatient pulmonary rehabilitation program. The characteristics of the two groups were similar. Their usual medication included inhaled beta agonists (n = 15), inhaled anticholinergics (n = 15), inhaled corticosteroids (n = 12), theophylline (n = 6), and oral corticosteroids (n = 4). At baseline, the mean FEV₁ was 602 (SE, 41) ml, and this increased by 113 (17.2) ml after nebulized salbutamol. The slow VC was 1,905 (123) ml at baseline and this increased by 302 (65) ml with salbutamol. The mean PO₂ was 74 (SE, 1.5) mm Hg (9.8 kPa) and the mean PCO₂ was 40 (SE, 0.75) mm Hg (5.3 kPa). During the run-in the mean daytime breathlessness score recorded on the diary cards was 2.3 (0.15), and the nighttime score was 0.9 (0.22). At the beginning of the first treatment period the mean 6MW distance was 268 (27) meters. Two subjects withdrew before completing the first treatment period. Both were receiving morphine. All the results listed below, except adverse effects, refer to the remaining 14 patients.

Dose Achieved

At the end of the dose titration, the mean daily total dose was 2.5 (0.2) capsules for morphine (i.e., a total of 25 mg sustained-release morphine sulphate) and 3.4 (0.3) capsules for placebo. Compliance, determined by tablet counts, was in excess of 90% for all patients.

CRQ

The changes from baseline in the CRQ scores are shown in Table 1. More positive scores are favorable. There was no difference between morphine and placebo in the total score for the CRQ. The mean changes from baseline for the total score were 2.08 with morphine and 2.94 with placebo (p = 0.95). The minimum clinically important difference (MCID) for the total CRQ score has been estimated as 10 points (21). On the Mastery subscale subjects fared worse during treatment with morphine than with placebo (p = 0.02). For the Mastery subscale, the MCID is 2 points. With placebo, the mean improvement on the Mastery subscale was 2.51 points, whereas with morphine, there was a decrease of 0.36 points. There was no significant difference between the two treatments on the Dyspnea, Fatigue, and Emotional subscales.

Six-Minute Walk

On six occasions during the study the subjects were unable to undertake the 6MW (For two subjects this was on one occasion, and for two subjects it was on two occasions). When this was due to breathlessness (four occasions) the subjects were assigned a distance of 0 meters. If they were unable to walk because of another physical complaint (sore back or foot) they were assigned the distance walked at the beginning of the first treatment period.

The 6MW distance decreased by 35.1 (18.9) meters between the beginning and the end of the morphine treatment period. In contrast, the distance walked increased by 21.6 (16.6) meters during the placebo treatment. There was no evidence of an order effect, and the distance walked was significantly worse (p = 0.04) with morphine.

If the subjects who were unable to walk because of breathlessness are excluded, the distance walked was still 15.1 m less with morphine and 14.2 m more with placebo (NS).

At the end of the morphine treatment period, the breathlessness scores on the Likert scale were 5.4 before the 6MW and 2.0 at the end of the 6MW. This compared with values of 5.7 and 2.3 for the 6MW at the end of the placebo treatment period. There was no significant difference in 6MW breathlessness scores between the two treatment periods.

Diary Scores

There was no difference between treatments in breathlessness scored at home twice daily. The mean daytime score with morphine was 2.22 and with placebo it was 2.26 (possible scores, 0 to 5). The mean nighttime score was 0.81 with morphine and 0.85 with placebo (possible, 0 to 4).

Spirometry and Oxygen Saturation

During treatment with morphine FEV₁ and VC decreased by 32 (26) ml and 81 (69) ml, respectively. With placebo, FEV₁ decreased by 67 (27) ml and VC increased by 14 (114) ml. None of these changes was significant.

Baseline oxygen saturations at rest were the same (95%; SE, 0.44%) at the beginning of both treatment periods. The mean change in saturation from baseline was +0.28% (0.41) with morphine and +0.13% (0.25) with placebo. No subject dropped their saturation more than 4% between readings, and on no occasion was the saturation at rest less than 91%.

Adverse Effects

Adverse effects are listed in Table 2. At each visit subjects were specifically interrogated about nausea, anorexia, constipation, or drowsiness. These are the adverse effects that one would anticipate with morphine. Adverse effects were also recorded if they were reported in response to the question "have you experienced any other side effects since the last visit?" or if they had been recorded on the diary card. There was no difference in the overall incidence of adverse effects (p = 0.25), but subjects were significantly more likely to report nausea, anorexia, constipation, or drowsiness (p = 0.004) while receiving morphine.

Two subjects withdrew from the study while receiving morphine (both during the first treatment period), one after 10 d with a severe infective exacerbation of COPD and the other after 21 d with severe constipation and shakiness secondary to the medication (20 mg twice a day).

An opiate withdrawal syndrome was observed in three patients when treatment was stopped at the end of the first 6-wk treatment period, and in the subject who withdrew after 21 d. This consisted of sympathetic overactivity, insomnia, diarrhea, generalized lassitude, and increased breathlessness and sputum production. Two of these four patients required treatment with oral steroids, and one had a clonidine patch administered to reduce withdrawal symptoms. Most symptoms resolved in 1 to 2 wk, although sleep disturbances persisted for longer. All subjects who completed the first treatment period completed the study.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we did not demonstrate any benefit of treatment with sustained-release morphine. The Total CRQ scores were similar for morphine and placebo, but on the Mastery subscale there was a significant difference, which favored placebo. With morphine there was decrease of 0.36 points, but the placebo there was an improvement of 2.51 points (compared with a MCID of 2 points). This suggests that overall, patients feel less control over their disease with morphine than with placebo. The mean improvement in breathlessness on the Dyspnea subscale was 2.5 points for morphine compared with only 0.44 points for placebo, but this difference was not statistically significant. This could represent a type II error, but morphine did not result in any improvement over placebo on breathlessness measured on the daily diary cards or at the end of the 6MW, so it is unlikely that we have missed a significant symptomatic benefit with this treatment. Indeed, morphine appears to be deleterious. With morphine, not only did the subjects fare worse on the Mastery subscale, but they did not walk as far in 6 min, and the side effects of nausea, constipation, and drowsiness were more common.

We did not see any significant change in oxygen saturation at rest during treatment with morphine. Oxygen saturation increased by a mean 0.3% with morphine and 0.1% with placebo. This is in contrast to the study by Light and coworkers (7) in which a single dose of 0.8 mg/kg of morphine led to an improvement in exercise tolerance, but the Pa_O2 at the end of exercise was reduced from 71.9 mm Hg with placebo to 65.8 mm Hg with morphine. They concluded that a reduction in ventilatory drive was contributing to the improvement in exercise tolerance. The dose of morphine that we used did not reduce oxygen saturation and therefore may not have been high enough to reduce ventilatory drive. Alternatively, tolerance to the effects of morphine on ventilatory drive may develop with chronic dosing.

We could have administered a higher dose of morphine, but it is unlikely that this would have been tolerated. The dose of morphine was increased stepwise over 2 wk to try and minimize unacceptable adverse effects. Despite this, the average dose of morphine was only 25 mg/d compared with a target of 40 mg/d. Even with this dose, side effects were more common with morphine, and the lower score on the Mastery subscale of the CRQ may reflect this. We also observed an opiate withdrawal effect in several patients after a relatively short period of treatment with morphine.

Our findings are similar to those of Eiser and coworkers (11). They also failed to demonstrate an improvement in exercise tolerance or in breathlessness on diary cards or after a 6MW. They did not measure quality of life, although they reported that some patients complained of nausea and three patients had vomiting or constipation. Adverse effects were also prominent in a study by Woodcock and coworkers (10). With 30 mg three times a day of codeine they reported a reduction in subjective disability as measured with the oxygen cost diagram, but five of the 16 subjects withdrew because of nausea and vomiting.

Morphine-related adverse effects were sometimes very apparent, and the study would have been strengthened if the investigator who conducted the 6MW and administered the CRQ was blinded to these adverse effects. Nonetheless, we doubt that the result of the study would have been different because our original bias had been that morphine would be effective in reducing breathlessness. It is unlikely that the failure to blind this investigator to adverse effects led us to underestimate any potential benefit from morphine.

Nine patients elected to have open treatment with morphine therapy at the end of the double-blind phase, although only four continued to receive it after 3 mo. Subjects and investigators were unaware of initial treatment allocations during this 3-mo open phase. Subjects who chose open-phase treatment tended to fare better on morphine during the double-blind phase than did the other subjects (but the difference was not significant). We did not enquire as to why subjects chose to continue into the open phase, but we suspect that they enjoyed having regular clinical review, or hoped that they might yet improve. Morphine was usually discontinued early during the open phase, however, because of side effects and perceived lack of benefit. One patient, however, had spectacular success with MST 10 mg twice daily with a marked improvement in breathlessness and exercise tolerance. She continues to receive MST after 2 yr. Robin and Burke (22) reported a single patient who did well with hydromorphone in an n-of-1 clinical trial (23). We cannot exclude the possibility that there is a small subset of patients who do well when they are given morphine for breathlessness, but we could not recommend it as treatment for most patients with severe breathlessness caused by COPD.