pain and gabapentinoids: no clear benefit, potential harm

 Yet another article found that gabapentinoids (gabapentin and pregabalin) were ineffective for pain control, this one finding no benefit in pain reduction or improved sleep in patients with peripheral nerve injury (see gabapentinoid not reduce periph nerve pain ClinOrthopRelatRes2025 in dropbox, or DOI 10.1097/CORR.0000000000003415. Thanks to orthopedic surgeons Dickey Jones and Brad Henley for bringing this to my attention

 

Details:

-- there were only four randomized controlled trials included in this meta-analysis on gabapentinoids versus placebo for patients having nerve injuries, in a literature search from 2000 to 2022

    -- these four studies involve 919 patients: 402 were treated with either gabapentin or pregabalin (gabapentinoids), 394 were treated with placebo, and 123 with both in the two crossover trials

        -- the researchers assessed only the initial blinded side of the crossover trials, since the adverse effects of gabapentinoids (specifically dizziness, somnolence, and gait disturbance) would reveal to the patient whether they were then put on an active medication versus a placebo. So, the initial component of the study, when the patients were blinded to active medication versus placebo, was the only part included

-- in the three of the fourth studies having more granular data, women represented 57% of those on treatment and 47% on placebo; mean age was 52

-- risk of bias using the Cochrane Risk of Bias tool, was very low for all of the included studies

-- none of the studies had objective consensus reference tests to diagnose peripheral nerve injury, but relied on clinical judgment

    -- the pain intensity was measured by either a visual analog scale (VAS) or a numeric rating scale

 -- three studies assessed pregabalin doses ranging from 150 mg to 600 mg, and one of gabapentin from 300 mg to 2400 mg

 

-- the minimum clinically important differences (MCIDs) were as follows:

    -- for pain: in the medical literature, the range on a VAS scale from 1 to 10 (the higher the score, the worse) determined the MCID to be 1.6 to 3.0; in this study they choose a 2.0 difference

    -- for sleep: on an 11-point score measuring how pain affected sleep (the higher the score, the worse), there is not much information in the medical literature, but one study concluded that a one or two point change would be clinically important; this study considered a 1.5 change to be the clinically important difference for the daily sleep interference

 

--  main outcomes:

    -- effect of gabapentinoids versus placebo in reducing pain at one month as well as at 2 to 4 months after treatment

    -- effect of gabapentinoids versus placebo on sleep interference at 2 to 4 months

 

Results:

-- pain at one month:

    -- gabapentinoids versus placebo: -0.21 (-0.72 to 0.29), P=0.40; not statistically significant

-- pain at 2-4 months:

    -- gabapentinoids versus placebo: -0.38 (-0.76 to 0.00), P=0.05, statistically significant

        -- however, this level of change did not reach the criteria above for being the clinically important difference of -2.0

 

-- sleep interference at 2-4 months:

    -- gabapentinoids versus placebo: -0.56 (-0.91 to -0.22), P<0.01, statistically significant

        -- however, this level of change was not close to the clinically important difference of -1.5

 

Commentary:

-- peripheral nerve injury is not uncommon, about 2.8% of all patients reporting to a level I trauma center have such an injury, and about half of them experience substantial pain more than six months after that injury

-- there is a not surprising relationship between pain and sleep, which is bidirectional: pain can cause poor sleep quality in about 80% of patients, and poor sleep quality can lead to reduced pain tolerance

-- gabapentinoids have been prescribed increasingly to reduce the intensity of peripheral nerve injury-related pain and improve sleep; however, this indication is off-label

    -- that being said, we clinicians do prescribe medications off-label quite often: drug companies tend to seek FDA approval based on designing the initial approval study that is most likely to show medication benefit

        -- and they have basically no financial interest in doing further studies after the FDA approval that could support other indications, since these later trials might find that the medication was ineffective. such results could significantly decrease the number of prescriptions being written and affect their bottom line...

     -- and this issue is particularly important given the high frequency of adverse effects from gabapentinoids (see below)

         -- which means that there really should be some mechanism to require further studies to document benefit for the non-targeted initial conditions, since clinicians often will prescribe medications that they think are likely to be effective for the patient.

         -- another and much more pervasive issue is the inherent limitations of pretty much all studies: they typically exclude patients who have chronic kidney disease, or other medical conditions, or are older than the age limit defined by the study, or from a different area of the world with different culture/healthcare systems/diet and exercise/overall environment, social conditions, support systems, etc. However, when we see such patients who would not have been included in the study, we typically still prescribe the targeted medications if we feel it is necessary/appropriate (ie, even "good quality" studies do not apply to many of the patients we see daily)

            -- so, this also brings up the fact that we are prescribing medications without clear guidance from well-designed studies, and these medications may well have significant adverse consequences

        -- that being said, in 2022 gabapentinoids were the 10th most prescribed drugs in the United States, with <1% prescribed for FDA indications; about 5% of off-label usages were for depressive disorders, 4% for anxiety disorders, and 2% for bipolar disorders. More than half of the off-label prescriptions for patients included prescribing other CNS medications, specifically antidepressants, opioids, and benzodiazepines (https://pubmed.ncbi.nlm.nih.gov/34015964/ )

-- and, of note, there were significant drug company shenanigans/malfeasance of inflating the drugs effectiveness in treating non-neuropathy-related pain, leading to the manufacturers of both gabapentin and pregabalin having to pay penalties for improper marketing: https://www.nejm.org/doi/full/10.1056/NEJMp0808659 and https://www.nejm.org/doi/full/10.1056/NEJMsa0906126 , the former article revealing blatant lies and exaggerations, specifically noting that pain management is where the money was at and gabapentin was good for "everything" including bipolar disease  and all pain

 

-- this study found no significant benefit in gabapentinoids versus placebo in reduction in pain or sleep disruption after nerve injury. Though this was a small study, it fit in with several other studies suggesting that gabapentinoids were not very effective for chronic pain, with the possible exceptions of postherpetic neuralgia and diabetic neuropathy:

    -- though there are some differences depending on how the systematic review/meta-analysis was done, a meta-analysis of 9 studies found that gabapentinoids did not help with low back or radicular pain: https://gmodestmedblogs.blogspot.com/2018/07/gabapentinoids-still-not-help-low-back.html  

    -- an RCT found that gabapentin was ineffective for chronic pelvic pain in women: https://gmodestmedblogs.blogspot.com/2020/10/chronic-pelvic-pain-gabapentin.html  

    -- a systematic review of 21 RCTs of pregabalin or gabapentin vs placebo for diabetic peripheral neuropathy found either no effect (5 RCTs for gabapentin) or a small effect (16 RCTs, low strength of evidence, for pregabalin): see http://gmodestmedblogs.blogspot.com/2017/04/diabetic-peripheral-neuropathy-and-more.html , with much better results for other meds (eg duloxetine, venlafaxine, tricyclics)

-- there are clearly some patients who do find that gabapentinoids help them with pain.  What would be the explanation for this, given that the randomized controlled trials do not support their efficacy?? It turns out that this concern was in fact substantiated in the current study by the finding that there was “no difference in pain intensity favoring gabapentinoids before crossover, while there was a difference after crossover in both trials” per the researchers in this paper. So, more detailed analysis of the 2 crossover trials found no benefit for people who were initially randomized to a gabapentinoid vs placebo, but after crossing over to the opposite intervention and receiving the opposite trial drug, the patients did benefit from the gabapentinoid, presumably because they could tell if they were initially on the gabapentinoid or were switched to one, since they would more likely experience an adverse effect if on the “real” medicine, thereby enhancing the placebo effect.

    -- and there are also serious consequences of gabapentinoids http://gmodestmedblogs.blogspot.com/2017/04/diabetic-peripheral-neuropathy-and-more.html:

        -- those with COPD can have severe COPD exacerbations, presumably related to the gabapentinoids' respiratory depression: http://gmodestmedblogs.blogspot.com/2024/01/gabapentinoids-association-with-severe.html

        -- there has been a dramatic increase in gabapentin overdoses, and these are associated with respiratory depression as well as to gabapentinoids being meds with high misuse/diversion potential (see http://gmodestmedblogs.blogspot.com/2020/01/gabapentin-and-baclofen-overdoses.html ).

        -- there also may be a higher opioid death rate when opioids are used in combination with gabapentinoids (despite prior CDC suggestions to add them to opioids as an adjunctive therapy to decrease opioid use, see http://gmodestmedblogs.blogspot.com/2018/09/gabapentanoids-plus-opioids-higher.html ) 

            -- a recent study found that gabapentinoids were associated with a 2-to-3-fold hazard ratio for substance misuse and overdoses: https://pubmed.ncbi.nlm.nih.gov/38662459/ , especially in those with prior history of substance misuse disorders but also in those with psychiatric comorbidities

            -- another study found that 60% of patients had central nervous system (CNS) side effects including drowsiness, dizziness and ataxia, increasing the risk of falls and the attendant injuries and fractures: https://pmc.ncbi.nlm.nih.gov/articles/PMC11404328/

 

Limitations:

 -- there was not a lot of granular information in determining the minimum clinically important differences (particularly for sleep), so the researchers made assumptions from a few studies that dealt with patients with diabetic peripheral neuropathy and postherpetic neuralgia, a group very different from the one studied above

-- peripheral nerve injury was not assessed systematically by objective criteria, though all but one study reported confirmation of nerve injuries by pain specialists or neurologists

    -- this issue is an important one because peripheral nerve injuries are frequently associated with other soft tissue or bony injuries (e.g. fractures and lacerations), so the conclusions specifically regarding peripheral nerve injury as the cause of pain and sleep disturbance are not necessarily related to injury of the nerve. However, as noted above, several studies on gabapentinoids for chronic pain, and neuropathic pain are consistent with the finding of lack of benefit

-- the study combined gabapentin (in one study) and pregabalin (three studies) and with different dosages. There is evidence that pregabalin is more consistently absorbed and with more reliable systemic levels than gabapentin. However, this study did find no difference in their subgroup analysis, noting overlapping confidence intervals

-- the studies selected in this meta-analysis all excluded missing patients from their intention-to-treat analysis

-- and although there were 919 patients included in the meta-analysis, this comprised only 4 RCTs, perhaps limiting the reliability of the findings and limiting the ability to do sensitivity analyses for different scenarios

 

So, what is the bottom line here??

-- it seems to me that the clinical benefits of gabapentinoids in reducing pain (and, relatedly, improving sleep quality) are grossly exaggerated

-- and, it seems likely that whatever positive benefits from taking gabapentinoids in individual patients are likely from the placebo effect, triggered by that individual’s experience with the very common adverse effects of gabapentinoids (fatigue, dizziness, GI adverse effects…) and leading patients to believe that this is a strong drug and therefore likely to be a beneficial medication

-- and, as per above, some patients have really serious adverse effects, especially if they have some respiratory compromise (eg COPD exacerbations getting worse with gabapentinoids), or potentially with other medical conditions associated with respiratory issues (heart failure, chronic kidney disease), or other meds that are associated with respiratory compromise (sedatives, opiates, etc), and perhaps alcohol (https://www.ncbi.nlm.nih.gov/books/NBK557381/ )

    -- and, gabapentinoids are associated with significant adverse consequences; these are not medications that should be prescribed lightly and without clear scientific support...

-- the meta-analyses above did find that there are other meds that help with chronic pain, especially duloxetine and venlafaxine (and my own experience supports their benefits found in RCTs)

-- another option that may work incredibly well is topically diclofenac, which seems to be better than oral NSAIDs for pain treatment: https://gmodestmedblogs.blogspot.com/2020/08/acute-musculoskeletal-pain-topical.html

geoff

-----------------------------------

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

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

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