Data from a pooled analysis of three studies suggest that in patients with pain caused by diabetic nerve damage, or diabetic peripheral neuropathy, who are treated with Cymbalta (duloxetine HCl), improvements in both average daily pain and night pain severity were associated with less pain-related sleep interference than in those patients taking sugar pill. Results from the analysis of more than 1,000 patients were presented today at the annual meeting of the American Pain Society in Washington D.C.

The average daily night pain severity and sleep interference correlation assessments were secondary analyses of data from trials in which the primary efficacy endpoint was to measure the reduction in average weekly pain. For these assessments, a subset of nonsomnolent patients was created by removing those who experienced treatment-related somnolence (such as daytime sleepiness, drowsiness, grogginess or difficulty awakening) or who were on other sedating medications. At the end of the 12-week analysis, this subset of patients treated with 60 mg of Cymbalta once- and twice-daily experienced significantly more improvement in 24-hour average pain scores than those taking sugar pill (47 %, 50 %and 29 %respectively), as measured by reduction of scores on the self-reported, 11-point Likert scale.

In analyses of the effect of Cymbalta on night pain and pain-related sleep interference, nonsomnolent/nonsedating drug patients treated with 60 mg of Cymbalta once- and twice-daily experienced significantly more improvement in nighttime pain than those taking sugar pill after one week of treatment (22 percent, 21 percent and 10 percent, respectively) and at the end of the 12 weeks (47 percent, 51 percent compared with 34 percent). Both doses of Cymbalta reduced pain-related sleep interference at the end of the 12 weeks significantly more than sugar pill (55 percent, 57 percent and 45 percent respectively), as measured by the Brief Pain Inventory (BPI). No significant difference was seen between the 60 mg and 120 mg doses of Cymbalta.

Of greatest interest is the fact that for the first time, a clear correlation between reduction in average daily pain and average night pain and reduction in pain sleep interference was demonstrated in a population that was not depressed, did not experience treatment-emergent somnolence, and was not on medications commonly associated with sedation. Other published studies of the relationship between pain and sleep interference have been confounded by at least one of those conditions.

"Patients with diabetic nerve pain often complain of being awoken, or having difficulty falling asleep because of pain, and it is difficult for physicians to know whether sleep problems are the cause or effect of pain," said David Fishbain, M.D., distinguished professor of psychiatry, adjunct professor of neurological surgery and anaesthesiology at the University of Miami, and lead study author. "This new analysis is important because it suggests that interference with sleep generally improves if nighttime pain improves."

In this analysis, the most common adverse events experienced in the entire patient population as a whole were nausea, somnolence, dizziness, constipation, hyperhidrosis, dry mouth and decreased appetite. The median duration of nausea, somnolence and dizziness in Cymbalta-treated patients\ was six days, 14 days and five days, respectively, with the majority of patients who experienced these adverse events rating them as mild to moderate in severity. The overall discontinuation rate for patients on Cymbalta was 14 percent, while the overall discontinuation rate for patients on placebo was 7 percent. The most common adverse event cited as a reason for discontinuation was nausea (4 percent), followed by somnolence (2 percent), dizziness (2 percent), fatigue (1 percent) and vomiting (1 percent).