Controlled-release oxycodone for pain in diabetic neuropathy

Subjects.

A total of 159 adults with a history of stable diabetes mellitus and a glycosylated hemoglobin (HbA_1C) level of ≤11% were randomized and received at least one dose of study medication.

Subjects had painful symmetrical distal polyneuropathy documented by neurologic evaluation using the Einstein Focused Neurologic Assessment (J.C. Arezzo and H.H. Schaumburg, written communication, April 1999). Investigators trained in this procedure assigned an overall clinical impression scored on a four-point categorical scale to assessments of distal sensory function, distal muscle strength, and select reflexes. Distal sensory function was examined bilaterally at three standardized sites on the lower limbs (base of great toe, anterior malleolus, and head of fibula) using the following stimuli and instruments: light touch (Semmes-Weinstein monofilaments); temperature (cold only; metal thermal rod); and sharp (pin). Distal muscle strength (finger spread, great toe extension, ankle dorsiflexion) and deep tendon reflexes (biceps brachii, quadriceps femoris, and Achilles) also were assessed bilaterally.

Candidates for the study were required to have a pain syndrome consistent with painful symmetrical distal polyneuropathy; a history of pain in both feet (defined as an average pain intensity score of ≥5 on a numeric scale of 0 to 10) for more than half the day for at least 3 months prior to enrollment; and at least moderate pain in the absence of any opioid analgesic therapy for 3 days before receiving the study treatment. All prestudy opioid drugs were discontinued before administration of study medication.

Subjects were excluded from participation if they had 1) unstable or poorly controlled diabetes; 2) chronic pain unrelated to diabetic neuropathy; 3) a history of substance or alcohol abuse within the past 10 years; 4) serum creatinine levels ≥2.5 mg/dL; 5) hepatic dysfunction ≥3 times the upper limit of normal; 6) a history of active cancer, excluding basal cell carcinoma of the skin, in the past 3 years; 7) hypersensitivity to oxycodone or opioids; 8) rapidly escalating pain or recent neurologic deficit within the previous month; 9) a total of more than three doses per day of a short-acting opioid formulation in the preceding 2 weeks; 10) treatment with any long-acting opioid formulation; 11) autonomic neuropathy or gastrointestinal dysfunction that could compromise drug absorption or increase the risk from therapy; or 12) a need for elective surgery involving preoperative or postoperative analgesics or anesthetics during the study period. Women were excluded if they were pregnant or breast-feeding.

Treatment assignment.

Prior to study start, subjects were assigned to treatment using a randomization schedule with permuted blocks of size 4 which was generated by the sponsor with SAS software (SAS Institute, Cary, NC). The schedule was used to package and label the study medication shipped to the investigative sites. The randomized treatment assignment was executed by the site staff, who assigned subject numbers in ascending sequence as subjects qualified for randomization. Site staff dispensed the blinded medication. Sealed randomization information was provided to the sites if needed for an emergency. Sealed copies of the randomization schedule also were placed in locked sponsor files. Unblinding took place according to a standard procedure after the database was locked and the analysis plan, including evaluability determination, finalized.

Procedures.

Subjects were screened 3 to 7 days prior to entering the treatment phase. During the screening visit, a history was taken, subjects had a general physical and neurologic examination, and blood was sampled for HbA_1C, and renal and liver function tests. Subjects discontinued all opioid therapy at least 3 days before starting any study medication. Outpatient visits were scheduled on days 0 (start of treatment), 14, 28, and 42 (or at the time of discontinuation from study therapy) of the double-blind treatment phase.

Subjects received CR oxycodone (OxyContin; Purdue Pharma L.P., Stamford, CT) or an identical placebo tablet every 12 hours during the treatment phase of the study. Dosing began with one tablet of CR oxycodone (10 mg) or matching placebo. Subjects could increase the dose by one 10-mg tablet of CR oxycodone or placebo in the morning and one 10-mg tablet in the evening. Upward titration could occur every 3 days up to a maximum dose of 6 tablets (60 mg CR oxycodone) twice daily. No opioid rescue was allowed. If a subject developed unacceptable side effects, the dose could be decreased to the previous acceptable level. A medication compliance worksheet was completed at each study visit. Treatment lasted up to 6 weeks. After the end of the study, a final 1-week taper was optional.

Medications taken for diabetes control as well as adjuvant pain medications were continued at the same stable prestudy dose. Except for study medications, no other opioid analgesics were permitted. Other nonopioid analgesics, such as nonsteroidal anti-inflammatory drugs or acetaminophen, taken at a stable dose for ≥3 weeks before the study could be continued at the same prestudy dose. Stable dosing of concomitant medications was documented at each study visit.

Efficacy variables.

The primary efficacy analysis was based on average daily pain intensity during the past 24 hours obtained during the study period from days 28 to 42. Pain intensity was rated in a daily diary using an 11-point scale ranging from 0 (no pain) to 10 (pain as bad as you can imagine). During the screening phase, subjects completed the diary for a minimum of 3 days without opioid therapy and qualified for the study if they attained an average pain intensity score of ≥5. The daily diary also included 0 to 10 scales for current pain and worst pain; a 1 (not satisfied) to 6 (totally satisfied) scale for satisfaction with pain medication; and a 0 (poor sleep) to 10 (excellent sleep) scale for sleep quality. All scales were numeric.

Secondary end points included 1) scores recorded in the daily diary for average pain intensity from days 1 to 27, and for current and worst pain, satisfaction with pain medication, and sleep quality from days 1 to 42; 2) total and subscale scores calculated from the 14-item Brief Pain Inventory (BPI)29 administered on days 0, 14, 28, and 42, or at study discontinuation; 3) scores for validated measures of psychological state (Rand Mental Health Inventory30), physical functioning (Sickness Impact Profile31), and general health status (SF-36 Health Survey32) obtained on days 0 and 42, or at discontinuation; 4) proportion of subjects who discontinued the study medication due to lack of efficacy; and 5) time to mild pain, number of days with mild pain, and proportion of days with mild pain.

Safety variables.

The investigator asked about adverse events at each study visit, and reported events were graded for severity and probability of relationship to the study drug.

Statistical analyses.

Based on an expected SD of 2.5 for the pain intensity score, 64 evaluable subjects per group were required to detect a difference between treatment groups of 1.25 on an 11-point scale with 80% power and a 5% level of significance. Planned enrollment was 160 subjects. The intent-to-treat (ITT) efficacy analysis included all subjects who received at least one dose of study medication. Because one subject was excluded due to double enrollment, the ITT efficacy analysis included 159 subjects. A per protocol cohort of 146 subjects was also defined; this cohort excluded 10 subjects from the CR oxycodone group and 4 from the placebo group because of protocol violations. Missing observations due to discontinuation were extrapolated by carrying forward values from the last observation immediately before discontinuation up to and including day 42. Intermittent missing data were not imputed.

The primary efficacy analysis used repeated measures analysis of covariance (ANCOVA) with day as the repeated factor to compare the average pain intensity during the past 24 hours recorded in the daily diary (from days 28 to 42) between treatment groups. The period between days 28 and 42 was selected to allow for an initial titration period. The null hypothesis postulated no overall treatment group effect from days 28 to 42. Baseline pain, center, treatment, study day, age group, sex, treatment by day, and treatment by center were included as effects in the ANCOVA. Baseline pain was included as a covariate because it was considered a prognostic variable. Secondary efficacy analyses used the same statistical model but included baseline measurements for the respective variables instead of baseline pain. Results of the ANCOVA analyses are reported as least squares means, i.e., group means adjusted for covariates and other terms in the statistical model.

Time to mild pain was estimated using the Kaplan-Meier method, and the two treatments were compared by the log-rank test. The Wilcoxon test was used to compare treatment differences in the number and proportion of days with mild pain. Adverse events were classified by COSTART term and body system, and tabulated by treatment group. Fisher’s exact test was used to compare discontinuation rates and the incidence of adverse events between treatment groups. All statistical tests were two-tailed at the 0.05 level of significance.

Efficacy.

In the ITT cohort, the efficacy analysis of the primary end point showed that CR oxycodone provided superior analgesia to placebo (p = 0.002) (figure 2). Least squares mean scores (SE) for overall average daily pain intensity from days 28 to 42 were 4.1 (0.3) for the CR oxycodone group and 5.3 (0.3) for the placebo group on a scale of 0 to 10. The primary efficacy results from the per protocol cohort confirmed these results: least squares mean scores (SE) for overall average daily pain intensity from days 28 to 42 in this cohort were 4.2 (0.3) for the CR oxycodone group and 5.3 (0.3) for the placebo group (p = 0.009). In the ITT cohort, a significant difference in average daily pain intensity between groups was first apparent by day 3. These differences were maintained throughout the treatment period (see figure 2). The period of sustained analgesia from days 28 to 42 coincided with a mean average daily CR oxycodone dose of 42 mg.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Average pain intensity scores recorded in subject daily diaries from days 28 to 42 were reduced with the use of CR oxycodone (4.1 ± 0.3) compared with placebo (5.3 ± 0.3) (p = 0.002) in the ITT cohort. Scale ranges from 0 (none) to 10 (pain as bad as you can imagine). Values are expressed as least squares means (SE). = Placebo; = CR oxycodone.

Analyses of the secondary efficacy variables obtained from the daily diary in the ITT cohort yielded similar results. CR oxycodone produced significant improvements in overall scores for average pain intensity from days 1 to 27 and for pain right now, worst pain, satisfaction with study medication, and sleep quality from days 1 to 42 (table 3). Significant improvements in all pain measurements (except worst pain) and in sleep quality were observed within 1 week of initiation of CR oxycodone therapy.

A comparison of BPI scores obtained at baseline and at the end of the study showed that 9 of 14 items (average pain intensity, pain right now, worst pain, least pain, pain relief, interference score, relations with other people, sleep, and enjoyment of life) were significant and improved in the CR oxycodone group compared with the placebo group (table 4). On the pain relief scale of 0% (no relief) to 100% (complete relief), mean scores at the final study visit were 53% for the CR oxycodone group compared with 37% for the placebo group (p = 0.005). No significant improvements occurred for the five remaining items of the BPI (physical function score, general activity, mood, walking ability, and normal work).

The median time to achieve mild pain (defined as an average pain intensity score of ≤4 on the 0 to 10 scale) was shorter for the CR oxycodone group (6 days) than for the placebo group (17 days) (p = 0.017). In addition, subjects treated with CR oxycodone had more days with mild pain: mean (SD) of 20.0 (16.6) days vs 12.5 (16.0) days for the placebo group (p = 0.007). Subjects who received CR oxycodone reported a higher mean (±SD) percentage of days with mild pain (47% ± 39%) compared with placebo-treated subjects (29% ± 37%) (p = 0.006).

No significant differences were observed between CR oxycodone and placebo groups in the physical functioning, general health, and mental health subscales of the SF-36 Health Survey or in the seven subscales of the Rand Mental Health Inventory. A significant difference in ambulation, a subscale of the Sickness Impact Profile, was observed between CR oxycodone and placebo groups at the final visit; however, the other 15 subscales of the Sickness Impact Profile showed no significant differences.

Safety.

Safety was evaluated for all 159 randomized subjects. The incidence of adverse events was greater in the CR oxycodone group than in the placebo group. Overall, 80 (96%) of 82 subjects treated with CR oxycodone and 52 (68%) of 77 subjects who received placebo reported adverse events. Seventy-one (86%) of 82 subjects given CR oxycodone compared with 33 (43%) of 77 subjects given placebo reported adverse events that were considered related to the study drug by the investigator. The majority of adverse events in both groups were mild or moderate.

The most commonly reported adverse events in the CR oxycodone group were those usually associated with opioid use: constipation (35 subjects), somnolence (33 subjects), nausea (30 subjects), and dizziness (26 subjects) (table 5). The most common adverse events reported among placebo-treated subjects were headache (18 subjects), constipation (11 subjects), and dizziness (8 subjects). The most common adverse event leading to discontinuation was nausea (2 CR oxycodone and 1 placebo). Overall, 26 (32%) of 82 subjects treated with CR oxycodone and 10 (13%) of 77 subjects given placebo had adverse events that were dose limiting (i.e., dose reduced, dose interrupted, or dose discontinued).

During the study, one death occurred in the CR oxycodone group during the fifth week of dosing; the cause of death was acute renal failure unrelated to the study drug. Thirteen other serious adverse events were reported: four in the CR oxycodone group (flulike syndrome, hyperglycemia, hypoglycemia, and epistaxis) and nine in the placebo group (alcohol intoxication, ascites, chest pain, asthenia, nausea, diarrhea, vomiting, decreased body weight, and dyspnea). None of these adverse events was considered related to the study drug.