Wednesday, May 2, 2007
Long-term opioids for non-malignant pain; Suffering in advanced cancer
1)
The Clinical Journal of Pain has a study on the long-term use of opioids for chronic non-malignant pain (CNMP). The study looked at 230 patients taking extended release oxycodone after completing short term, controlled trials for CNMP (DJD, low back pain, diabetic neuropathy) - they agreed to be observed for up to 3 years while taking the oxycodone. The patients were 90% white, 57% female, their average pain was 5/10, mean oxycodone dose 30mg a day. The study was terminated early unfortunately due to 'administrative reasons' with the registry database so the data are a little opaque. This study presents almost an overwhelming amount of data but these are some of the more relevant ones to me:
A) 133 patients dropped out prematurely; 30% of these stopped oxycodone due to adverse effects and 14% stopped because of lack of efficacy (the remainder were lost to follow up or were discontinued from the registry for unclear reasons). Most of the patients who had stopped due to adverse effects had done so by 6 months. 64%, 39%, and 18% of the patients took oxycodone for a full 1, 2, and 3 years, respectively (remember some of this data is tainted by the early termination of the registry).
B) 44% of patients increased their oxycodone in the first 4 months; 24% increased it between months 4-6. After that almost no patients who continued on oxycodone needed dose increases (only ~10% a month). Across all time periods the average dose was ~50mg a day.
C) Pain decreased a little in the first few months. There were 17 patients whose oxycodone doses escalated without a concomitant decrease (some even had an increase) in pain.
D) Adverse events were relatively rare (but remember people were dropping out, some of whom were dropping out from adverse events, and data were no longer gathered on them). Interestingly constipation diminished over time. This is counter to other experience which suggests constipation never attenuates. On the other hand it may have been getting better because it was being treated better (there's no way of knowing this) or that those with the worst constipation were dropping out.
E) Worrying behaviors for substance abuse/addiction were very rare--2.6% of patients demonstrated this (these data were collected prospectively as part of the registry data and were based on the investigating physician recording them as part of the routine data collection process).
Given the problems with the early discontinuation of the registry its tough to know how to interpret these findings. They are more or less consistent with the dominant thinking about opioids for CNMP: most patients don't tolerate or find them useful for the long term (years), however the substantial minority that do can use them safely without significant side effects, never-ending dose escalation, or risk of addiction. A major caveat here is that these patients are a relatively select group: they could pass the screening rigors of getting into a controlled clinical trial and there is some reason to believe that this subset of patients is different (among other things, more socioeconomically stable) than the population at large.
After reading this paper, and given the ongoing concerns and controversies about long term opioid use for CNMP, one gets a sense of just how valuable a 'registry' could be for researching this issue and I hope that these researchers or others are able to create a new one.
There are two accompanying editorials/commentaries--here & here. The first in particular is a worthwhile read--it summarizes opioid controversies for a few thousand years and succinctly outlines what we know, what we don't know, and an appropriate research agenda. It also talks about the ongoing epidemic of prescription opioid abuse.
2)
Journal of Clinical Oncology has a study on suffering in patients with advanced cancer. It looked at 380 hospitalized advanced cancer patients (median age 66y, median survival 60 days, all were on palliative care units or had been seen by palliative care teams) in Canada who underwent semi-structured and -quantitative interviews about their experience including if they felt they were suffering, why they were, in what way, and how would they describe it. Half the patients didn't describe themselves as suffering and 25% more described it as mild only. 8% described extreme suffering. This study too presented a lot of data and I've written too much already today. To be brief: younger age and higher socioeconomic status correlated with increased reports of suffering, as did functional status (although time to death was equal in both groups). A lot of the things one would expect predicted suffering did: increased existential and social distress, increased anxiety and depressive symptoms, and increased physical symptoms particularly weakness and pain. In regression analysis physical symptoms retained the most independent predictors of reported suffering.
While there's tons of interesting stuff here, what's most interesting to me is the finding that 3/4 of the patients reported no or mild suffering only--these were people 2 months from death. Why is this? I don't know. It seems low to me and may be due to patient selection (these were patients who felt well enough to participate in a lengthy interview among other things). It does have a certain consonance with the recent bereavement study in which acceptance was the most common, immediate reaction to a loved one's death.
(The same issue also has a systematic review of pain management of hospitalized cancer patients--looking at trials of interventions to improve pain management. You can easily guess the results.)
3)
And finally--palliative care was mentioned recently in USA Today. It's a very positive, and only a little muddled, article. USA Today--I guess this means we really are going mainstream in the US. (Thanks to David Weissman for letting me know about this.)