Saturday, December 10, 2005
Spine has published a randomized trial of transdermal fentanyl vs. SR morphine for chronic low back pain. I immediately took note because here was a chance to evaluate these two drugs head to head. The study, unfortunately, is rife with problems, but there are some interesting findings nevertheless. It is a randomized, but non-blinded, study of transdermal fentanyl vs. SR morphine in "strong-opioid" naive patients with chronic low back pain. Basically they randomized ~670 patients to either 25mcg/hr of fentanyl or 30mg bid of MS Contin, and gave both groups reasonable, patient-directed titration instructions. Concern about starting long-acting opioids on opioid-naive patients was not mentioned, & to their credit they didn't have any major adverse events (given the recent excitement about this topic, one would have appreciate more mention of this from the authors, especially about how much "weak" opioid use--were these patients truly opioid naive, etc?)
The authors found that after an average of ~240 days, pain was the same & dosing was roughly equianalgesic between the groups (mean morphine dose of 140mg/day vs. fentanyl dose of 57mcg/hr). Pain ratings were equal in each groups (mean visual analog scores ~55/100mm) at the end of the study. I was perplexed that baseline VAS scores weren't mentioned at all; for that matter neither was baseline medication use; nor, while we're at it, were standardized as-needed short-acting analgesics part of the protocol. The abstract annoyingly makes the claims that while the two drugs were equivalent on the VAS, fentanyl was 'significantly' better for pain at rest and pain at night. All of this was based on a per-protocal analysis and the actual difference in percent of patients reporting pain at rest/night was miniscule (a few percentage points); my point here is that this is a likely meaningless finding that has made its way into probably the most important part of the article--the abstract. In fact, this was one of those articles where I had to read each paragraph twice to keep track of when they were reporting an intention-to-treat analysis or a per-protocol. This is all a bad sign. In addition, in the "Table 1" (which is always patient characteristics) they inexplicably divide the patients' pain into 'neuropathic,' 'nociceptive,' 'neuropathic with psychologic factors,' & 'nociceptive with psychologic factors.' Huh? There's no mention as to how these distinctions were made or by whom; nor did they do anything with these distinctions later on. This too is a bad sign.
What did truly seem to be significantly different between the groups was that constipation occured more frequently with morphine than transdermal fentanyl: 48 vs 31%. Frustratingly, starting or at least offering stool softeners/stimulant laxatives to the patients wasn't part of the protocol, despite it being a well accepted standard of care (although I'll be the first to admit that that's not data-driven). Nevertheless, this is the first data I've seen in a head-to-head trial supporting the notion that transdermal fentanyl is less constipating than oral morphine, not that I've ever looked.
Janssen, who has recently put out a press release warning about starting opioid naive patients on Duragesic, funded the study.