painful diabetic neuropathy meds

 A guideline was recently published on pharmacologic treatments for painful diabetic neuropathy by the American Academy of Neurology, updating their 2011 guideline (see dm neurop guidelines AAN2022 in dropbox, or https://n.neurology.org/content/neurology/98/1/31.full.pdf 

Details:

-- literature search from January 2008 to April 2020; 65 studies included

   -- all were RCTs with more than 20 participants, with prespecified 5 oral med classes: gabapentinoids, SNRIs, sodium channel blockers, tricyclic antidepressants and SNRI/opioid meds

-- the guideline committee considered the standard mean difference (SMD) of effects of medications over placebo, with an absolute value of 0.2, 0.5, and 0.8 as thresholds for small, medium, and large effect sizes

-- outcome: comparison of the different meds for symptom response in those with painful diabetic neuropathy (PND) by SMD

 

Results:

-- gabapentinoids (gabapentin and pregabalin): standard mean difference (SMD) 0.44 (0.21-0.67)

    -- pregabalin: SMD 0.29 (0.13-0.45), small effect, low confidence

    -- gabapentin: SMD 0.53 (0.22-0.84), medium effect, moderate confidence

-- serotonin-norepinephrine reuptake inhibitors (SNRIs: duloxetine, venlafaxine, desvenlafaxine): SMD 0.47 (0.34-0.60)

    -- duloxetine: SMD 0.50 (0.26-0.74), moderate effect, moderate confidence

    -- desvenlafaxine: SMD 0.25 (0.07-0.43), small effect, low confidence

    -- venlafaxine: no comment, though one study was identified

-- sodium channel blockers (carbamazepine, oxcarbazepine, lamotrigine, valproic acid, lacosamide): SMD 0.56 (0.25-0.87)

    -- valproic acid: SMD 0.86 (0.38-1.33), large effect, low confidence

    -- others: no comment

-- SNRI/opioid dual mechanism agents (tramadol, tapentadol): SMD 0.62 (0.38-0.86)

    -- tapentadol: SMD 0.78 (0.54-1.03), medium effect, low confidence

    -- tramadol: no comment

-- tricyclic antidepressants (TCAs: eg amitriptyline, nortriptyline, imipramine): SMD 0.95 (0.15-1.8)

    -- amitriptyline: SMD 0.95 (0.15-1.8), large effect, low confidence

    -- others: no comment

-- other medications assessed:

    -- ginkgo biloba: SMD 0.83 (0.48-1.18), large effect, low confidence

    -- nutmeg extract: no more effective than placebo

    -- buprenorphine transdermal patches: not clearly more effective than placebo

    -- capsaicin: SMD 0.30 (0.14-0.47)

    -- glyceryl trinitrate spray: SMD 1.19 (0.55-1.83), 1 study, medium effect, low confidence

    -- citrullis colocynthis is possibly more likely than placebo to improve pain, SMD 0.91 (0.36-1.45), large effect, low confidence

though, note that there are overlapping confidence intervals for all of these meds (ie, none is clearly preferable)

-- In direct comparison studies:

    -- pregabalin is probably more effective than carbamazepine, SMD 0.86 (0.50-1.21), large effect, moderate confidence

    -- venlafaxine is probably no more likely than carbamazepine to improve pain

    -- pregabalin is more likely than venlafaxine to improve pain, SMD 0.84 (0.48-1.20), large effect, moderate confidence

    -- amitriptyline is no more likely than gabapentin to improve pain

    -- duloxetine is possibly more likely than nortriptyline to improve pain, SMD 1.64 (0.63-2.65), large effect, low confidence

-- No significant differences by age or sex in pain reduction

 

Recommendations (all were level B in quality, reflecting the limitations of the underlying studies):

-- clinician should assess diabetic patients for PDN as well as concurrent mood and sleep disorders

   -- patients should understand that complete pain relief may not be achievable

   -- both depression and sleep disorders (eg obstructive sleep apnea) can affect pain perception

-- patient should be offered a trial of TCAs, SNRIs, gabapentinoids, and/or sodium channel blockers to reduce pain (level B recommendation)

    -- they do comment that cognitive behavioral therapy may be effective (found in one small pilot study), and CBT as well as exercise and mindfulness have been found to be effective in other pain states,

    -- valproate should not be used in women of childbearing age who might become pregnant

-- if medication does not work or has adverse effects, it is reasonable to offer a medication from another class

    -- the typical treatment duration to get full effect is 12 weeks (range 4-16). so, consider the med ineffective if not significant response in 3 months

-- opioids should be avoided for the treatment of PDN

Commentary:

-- diabetes is the most common cause of peripheral neuropathy, accounting for 32%-53% of cases, with 16% of patients with diabetes developing painful diabetic neuropathy (PDN)

    -- PDN can hugely affect quality of life, from changing patient self-image (pain constantly reminding patients of a persistent disability) to sleep interference and the multitude of associated effects (poor memory, personality changes/irritability, decreased daytime function)

-- currently the most common prescriptions for diabetic neuropathy have been opioids, then gabapentin, pregabalin, duloxetine, amitriptyline, and venlafaxine [to me, this is almost precisely the reverse appropriate order: see below]

--Other recommendations and systematic reviews have come to different conclusions, largely because of which studies they choose to include. Some do not include older studies, suggesting that they have less validity and rigor as compared to the newer studies (which may or may not be true). Or other inclusion/exclusion criteria for the studies that they do include.

    -- for example,  the American Diabetes Association in their 2020 guidelines recommended pregabalin, duloxetine, or gabapentin as initial pharmacologic treatments for neuropathy. 

    -- on the other hand, a review in 2017 published in the journal Neurology commented that pregabalin and gabapentin were pretty useless, venlafaxine and duloxetine had the largest effect size, and tricyclics were just below them (see http://gmodestmedblogs.blogspot.com/2017/04/diabetic-peripheral-neuropathy-and-more.html

        -- and, even low back pain or painful radiculopathy do not seem to respond to the gabapentinoids: http://gmodestmedblogs.blogspot.com/2018/07/gabapentinoids-still-not-help-low-back.html 

-- a 2021 RCT compared pregabalin, duloxetine, nortriptyline, and mexiletine for 402 patients with cryptogenic sensory polyneuropathy, found that nortriptyline and duloxetine were efficacious (utility function of 0.81 and 0.80 respectively) but pregabalin and mexiletine trailed (utility function 0.68 and 0.58) and the quit rates were 38% for nortriptyline, 37% for duloxetine, 43% for pregabalin and 58% for mexiletine, thereby finding the highest utility overall was 0.52 for nortriptyline, 0.43 for duloxetine, 0.05 for pregabalin and 0.00 for mexiletine [pretty convincing that the positive effect was best for nortriptyline and duloxetine and the med intolerance was lower for them], see neurop meds nortrip best jaamneuro2021 in dropbox, or doi:10.1001/jamaneurol.2020.2590

-- But It should be emphasized that nonpharmacologic therapies do help with chronic pain overall, for example mindfulness and tai chi (see http://gmodestmedblogs.blogspot.com/2017/06/tai-chi-for-knee-oa-mindfulness-for.html ), or cognitive behavioral theapy (see http://gmodestmedblogs.blogspot.com/2017/04/home-based-cbt-for-low-back-pain.html ). the AAN recommendation above focused on meds, but did mention the nonpharmacologic approaches briefly (which really have been tested for chronic pain and not diabetic neuropathy). but these are reasonable to try in those with PDN, since they may well help the pain and may have other positive health effects (eg stress reduction). And, acupuncture might be considered, with a few studies suggesting some benefit (see https://pubmed.ncbi.nlm.nih.gov/33150711/ )

    -- another option not mentioned in the newe guidelines is the potential effect of gingko biloba and glyceryl trinitrate spray. few studies done, low confidence (which is true for many of the "real meds"). but the little they found was pretty impressive: medium to large effects.aybe these should be considered in those who do not respond to the traditional treatments????

--My approach to diabetic neuropathy, which has worked quite well over the past several decades, is to start with tricyclic antidepressants. In general, I find they work extremely well, typically at low doses, and at least 70% of the time lead to complete or near-complete pain relief. I basically use two of them: nortriptyline (typically 10 to 50 mg, usually 10 to 20 mg is a sufficient), or desipramine (typically 25 to 50 mg is sufficient). I choose the former if the patient has a significant sleep disorder, and the latter otherwise.

    -- one real concern I have is that desipramine is not approved by Medicare and by some other insurers. It is cheap, generic, and has way fewer adverse effects (perhaps most important is the anticholinergic effects, for which desipramine has the least of all TCAs, especially important in an elderly population with diabetes). 

    -- My experience and those of a brilliant neurologist I know have been that TCAs are interchangeable for neuropathy.

    -- and, there was one small study from 1992 finding that desipramine and amitriptyline (the latter having the most adverse effects of any TCA) had very similar effectiveness for PDN, with fluoxetine only helping in those with an underlying depression (i have not found any other TCA comparison studies...) (see https://www.nejm.org/doi/full/10.1056/nejm199205073261904), where the bottom two lines (best outcomes) were for amitriptyline and desipramine:

 

    -- in terms of opiates/SNRIs: there are some studies (including cited in this guideline) finding benefit from tramadol (low m and kappa opioid receptor affinity, but inhibits reuptake of norepinephrine and serotonin, and also inhibits GABA release and stimulates dopamine D2 receptors) and tapentadol

        -- but there are clear problems with using opioids for pain management (addiction, diversion, others at home getting them...) and not great studies (as with most other meds) documenting benefit; prior guidelines from the American Academy of Neurology (AAN) and the CDC recommend trying to avoid opiates.

        -- i still think tramadol might still have a role in patients disabled by their neuropathy, when other nonpharm and med treatments are unsuccessful, including trying combos of the more acceptable ones. I do review risks/benefits with patient, including the results of a 2019 study finding increased all-cause mortality with tramadol for osteoarthritis in a large observational database study in the UK: https://jamanetwork.com/journals/jama/fullarticle/2727448 ). Also, there are also issues of misuse of gabapentinoids at a similar rate to tramadol (eg see CMAJ January 14, 2019 191 (2) E47; DOI: https://doi.org/10.1503/cmaj.180599; or https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5573873/  )

       -- i do have a couple of patients on tramadol, who did not respond to the usual agents, and have had good effect. this is my last choice med, and I do it sparingly and cautiously. but there are patients who do respond to it, and it has improved their quality of life dramatically (largely by improving sleep by decreasing pain)

        -- it should also be pointed out that there are an array of studies finding that opiates are not better for chronic pain than other medications: see http://gmodestmedblogs.blogspot.com/2018/03/opioids-not-better-for-chronic.html and even for post-op pain (see http://gmodestmedblogs.blogspot.com/2019/01/restrictive-postop-opioids-fewer-given.html for post-op pain) 

Limitations:

-- as noted above, there are often large discrepancies between different systematic reviews or the articles chosen for guidelines. some of these are related to different time periods of studies they choose to include (and many guidelines eliminate older studies on older meds, which does NOT mean these meds are not as good as the current ones, and in fact they may be much better given that have a long known history and have survived the vicissitudes of time. and the studies included in the different guidelines do have different inclusion/exclusion criteria (which may reflect a bias on which meds the authors think are better, which also may be reflected in their research interests or other incentives, other biases....) 

-- it is not statistically rigorous to compare the findings of one RCT of drug A vs placebo to drug B vs placebo, and decide that drug A was better since its outcome was better than drug B. this is done in network metaanalyses when direct comparisons are insufficient or not available, but these comparisons are fraught: the patients treated may be different (psycho-socio-demographically, degree of neuropathic pain, medical comorbidities, measurement tools used, study techniques used such as amount of contact with the patients during the studies, differences in how the placebo group was treated...).  so comparing SMDs by network metaanalysis should be interpreted with caution. and, per the above guidelines, probably inappropriate for us to pay too much attention to the smaller differences (SMD)  between the different meds

so, another guideline. it is concerning that these clinical guidelines come to inconsistent results. we in primary care may well review a new guideline reasonably well and change practice based on that.  but the reality is that for several reasons mentioned above, these inconsistencies reflect different ways of including and excluding studies. this does bring up questions of conflicts of interest, the role the leadership of the committee plays, the variety of people involved in the guideline development (eg, is it all superspecialists who do not treat many people with easier-to-treat PDM, the ones we see and treat in primary care, and thereby have a skewed perception of medication responsiveness in the select group they see??). And with many guidelines, they eliminate older studies which have tried-and-true meds that have worked well, even though they have not had studies of the equivalent rigor of current ones. My assessment, based on my practice and reading many of the systematic reviews/metaananlyses is to start with a TCA if reasonable for the patient. then try venlafaxine or duloxetine as my second line. i do try to steer away from the gabapentinoids because of so many studies not finding benefit, having lots of adverse effects, and the potential for misuse alone or in combo with other drugs. and suggest strongly the role of nonpharmacologic interventions (exercise, midfullness, relaxation, acupuncture)

geoff

 

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@uphams.org

 

to get access to all of the blogs (2 options):

1. go to http://gmodestmedblogs.blogspot.com/ to see them in reverse chronological order

2. click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

3. or you can just click on the magnifying glass on top right, then  type in a name in the search box and get all the blogs with that name in them

 

or: go to https://www.bucommunitymedicine.org/ , a website from the Community Medicine section at Boston Medical Center.  This site does have a very searchable and accessible list of my blogs (though there have been a few that did not upload over the last year or two). but overall it is much easier to view blogs and displays more at a time.

 

 

please feel free to circulate this to others. also, if you send me their emails, i can add them to the list