Opioids not better for chronic back/hip/knee pain

---

From: Geoff A. Modest, M.D.
Sent: Wednesday, March 7, 2018 8:36 AM
To: Geoff A. Modest, M.D.
Subject: Opioids not better for chronic back/hip/knee pain

The just-published SPACE (Strategies for Prescribing Analgesics Comparative Effectiveness) trial found no benefit of opiates for chronic back or hip or knee osteoarthritis pain (see opiate not help chronic back hip knee pain JAMA2018 in dropbox, or doi:10.1001/jama.2018.0899 ).

Details:

--240 VA patients with moderate to severe chronic back pain or hip or knee osteoarthritis (OA) pain despite analgesic use were enrolled

--moderate to severe pain was defined as having pain nearly every day for >6 months, and score of 5 or more on the PEG scale (a scale from 0-10 reflecting pain intensity, and pain interference in enjoyment of life and general activity). 

--patients on long-term opioids or substance use disorder (SUD) were excluded. Notably, those with severe depression or PTSD were included. from 2013-2015

--mean age 58, 13% women, 88% white, 25% at least 4 yr post high-school education, 40% employed, 65% had low back pain/35% hip or knee OA, 25% current smoker, 3% hazardous alcohol use, 10% illicit drugs in past yr, 22% moderate depression, 9% moderate anxiety, 21% PTSD

--perceptions pre-randomization (similar between later assigned groups): opioid thought to be effective 7.7 (max=10), opioid safety 5.8, nonopioid effective 6.6, non-opioid safety 6.5

--preferences pre-randomization (but significant differences with later assigned group): no preference in 60% later assigned to opioids/43% nonopioid, prefer opioid in 21% later on opioids/37% nonopioid, prefer nonopioid in 20% each group

--randomized to (each had multiple medical options in 3 steps):

    --opioids: step 1: immediate-release morphine, oxycodone or hydrocodone/acetaminophen; step 2: morphine sustained-action or oxycodone sustained-action; step 3: transdermal fentanyl. maximum daily dose of 100 morphine-equivalents (MEs). if doses were titrated to 60 MEs without response, then rotation to another opioid within the step was considered before dosage escalation

    --non-opioids: first step acetaminophen or NSAID; step 2 adding oral adjuvants (tricyclics, gabapentin) or topical lidocaine; step 3: pregabalin, duloxetine, or tramadol

--follow-up visits were monthly with med adjustments until stable regimen achieved, then every 1-3 months

--98% completed the trial

--main outcome: pain-related function over 12 months, per the Brief Pain Inventory (BPI) interference scale; main secondary outcome: pain intensity, per the BPI severity scale (BPI scales: 0-10, higher score more pain or worse function). For both of these BPI scales, a 0.7 difference was considered clinically important

Results:

--despite the above titrations, most patients in the opioid group were on low to moderate doses of opiates:

    --in the opiate group, by 12-months: 20% were on 0 morphine equivalents/d, 42% of 1-20mg, 24% 20-50mg, 13% >50mg

    ​--in the nonopiate group, by 12-months: 89% were on 0 morphine equivalents/d, 10% of 1-20mg, 1% 20-50mg, 0% >50mg (up to 11% of patients were on tramadol)

--Pain-related function: no difference over the 12 months: BPI interference scale, 3.4 for opioid group and 3.3 for nonopioid, p=0.58

    ​-- >30% improvement in BPI interference score in 59% on opioids and 61% nonopioids

--pain intensity: significantly better in nonopioid group; 4.0 in opioid group but 3.5 in nonopioid group, difference of 0.5 (0.0-1.0), p=0.03

    ​-- >30% improvement in BPI severity score in 41% on opioids and 54% nonopioids (-25.6 to 0.0), p=0.05

--adverse medication-related symptoms: 1.8 symptoms in opioid group vs 0.9 in nonopioid group, difference of 0.9 (0.3-1.5), p=0.03

    --only anxiety was significantly better in those on opiates: baseline GAD score was 4.0 of 21 in the opioid group, decreasing to 2.5 at 12 months, but 4.0 and increasing to 4.2 in the nonopioid group (p=0.02). however, overall anxiety scores were low, and only 9% of patients had moderately severe anxiety

--subgroup analysis: no difference if primary symptom was back pain vs OA of knee/hip, no difference if adjust for smoking

Commentary:

--a recent meta-analysis of opiates in chronic back pain did find a small and likely clinically unimportant difference in symptoms in patients given higher vs lower opiate doses (see Abdel Shaheed C. JAMA Int Med 2016; 176(7):958), though other meta-analyses have not found benefit of dose escalation

--another really positive aspect of this study is that they did enroll patients with psych issues, which truly reflects what we often see in primary care practice

--they did not list the adverse outcomes specifically, but used a standard 19 medication-related symptoms, and added memory problems and sweating. but no indication of what adverse effects were most common nor how severe they were. 19% of the opioid group and 8% of the nonopioid group discontinued therapy, largest subgroup (n=9) in the opioid group was "side effects", and in the nonopioid group (n=8) was "lack of benefit"

--though the pain intensity scores were significantly favoring nonopiates, the difference of 0.5 is not considered clinically meaningful, BUT the % of patients achieving a >30% improvement was significantly higher with nonopiates

--how generalizable are these results?? a VA study does represent a special population: not many women, medical/psychosocial differences between VA users and general population, limits on how high they could increase opiate dose (though, notably even with their dose escalation approach, most were only on low dose. did these patients not push hard enough for higher doses???)

--there have been many blogs on opiates over the past years, which provide basically 2 different complementary conclusions:

    --opiates are not usually an effective pain med

        --opiates were not better than NSAIDs for acute pain in an ED study (see http://gmodestmedblogs.blogspot.com/2017/11/opioids-not-better-than-nsaids-for.html​ )

        --opiates were not more effective than NSAIDs in another ED study for acute low back pain (see http://gmodestmedblogs.blogspot.com/2015/10/opiates-for-acute-low-back-pain.html   )

        --NSAIDs seem to work as well for renal colic as opiates (see for example: Teichman J. N Engl J Med 2004; 350:684)

        --and, this study which found opiates not particularly helpful with several really common chronic pain conditions

    --once on opiates, it is really hard to stop

         --12th graders who were at low-risk of future subsequent substance use disorder but were prescribed opiates for a clinical indication were at 3-fold increased risk of SUD by age 23 (see http://gmodestmedblogs.blogspot.com/2015/10/prescribed-opioids-and-future.html)

        --elderly patients seen in the ED by high-prescribing opiate MDs had a 30% increased likelihood of being on opioids long-term, vs patients seen by low-prescribers but with similar clinical conditions (see http://gmodestmedblogs.blogspot.com/2017/02/opiate-prescribing-in-elderly-and.html  )

        --patients put on opiates post-surgery were more likely to misuse opiates long-term if they were on longer prescriptions post-op, more so than if were on shorter-term high-dose opiates (see http://gmodestmedblogs.blogspot.com/2018/01/post-op-surgery-opiates-and-subsequent.html )

       --and, a blog earlier this week demonstrated that 75% of patients on long-term opiates for chronic pain were voluntarily able to decrease their opiate dose by 50%, though still were on 150 morphine equivalents (see http://gmodestmedblogs.blogspot.com/2018/03/tapering-opioids-in-patients-with.html )

so, the most fundamental conclusion of this study, supplementing several other studies as above, is that we as clinicians should do everything we can to avoid initiating opiate therapy, since they are not so effective, really hard to stop, and have large potentially devastating individual and societal consequences.  All of the pain med prescribing guidelines stress using nonopioid meds first along with nonopioid adjuvants to help. the above study really reinforces this: nonopiates do seem to work even better than opiates overall

--a couple of other comments
    --there may be some groups of patients where opioids are necessary, either because of types of pain, or their responses to pain including differences in mu-receptor function (pain thresholds are really different from person to person), their social support systems, their overall social situations, their concurrent psych issues, etc).  it would be useful to know this specifically, though the imperative is still to try the nonopiate pathway first

    --there are also other adjuvants not addressed in this study. 

        --a recent blog found that gabapentinoids (gabapentin/pregabalin) were not helpful for low back pain (see http://gmodestmedblogs.blogspot.com/2017/08/gabapentinoids-not-indicated-for.html )

        --another found pregabalin did not help neuropathic pain (see http://gmodestmedblogs.blogspot.com/2017/04/pregabalin-did-not-help-sciatica.html  ).

        ​--I personally have found pretty good patient outcomes with venlafaxine in this regard, and also with tricyclics (esp desipramine and nortriptyline). 

        ​--but there are also potential benefits from yoga (see http://gmodestmedblogs.blogspot.com/2017/06/yoga-for-chronic-low-back-pain.html  , taichi or mindfulness (see http://gmodestmedblogs.blogspot.com/2016/06/tai-chi-for-knee-oa-mindfulness-for.html  ), home-based CBT (see http://gmodestmedblogs.blogspot.com/2017/04/home-based-cbt-for-low-back-pain.html  ) etc

to receive these blogs regularly, please contact me at gmodest@uphams.org