Evidence-Based Conclusions about Neuropathy Drugs

This very interesting article comes from Neurology Today (see link below) and looks at the effectiveness of certain drugs used to treat neuropathy, based on research and trials. The general conclusion is that Pregabalin fits the bill best but if you read the Editor of Neurology Today's own conclusions (below), it reflects not so much how wonderful Pregabalin is but the scattered and random nature of the studies and trials used to determine drug value. The fact is that there just haven't been enough studies of the various neuropathy treatments carried out; so that the one drug comes out better than the other because there is more available evidence concerning that particular drug and not because it necessarily compares better to all the others. Therefore Pregabalin becomes a Level A drug here because the research has been wider and better, than for other comparable drugs. This is not to say that Pregabalin isn't effective - for many people it clearly is but the comparisons with other drugs are just not sufficiently tested and evaluated. The only trustworthy conclusion to be drawn here is that once again, you shouldn't believe all you read in the papers!

(Don't be put off by the word 'Diabetic' in the title; the article applies to neuropathy in general.)

New Evidence-Based Guidelines for Painful Diabetic Neuropathy
Samson, Kurt: Neurology Today: 05 May 2011 - Volume 11 - Issue 9 - pp 1,4-5

ARTICLE IN BRIEF

The AAN, the American Association of Neuromuscular and Electrodiagnostic Medicine, and the American Academy of Physical Medicine and Rehabilitation jointly released a guideline, suggesting the best available evidence supported use of pregabalin for treating painful diabetic neuropathy.

A new evidence-based guideline for treating painful diabetic neuropathy (PDN) supports the use of several therapeutic options, most significantly the use of the seizure drug pregabalin, which was approved by the FDA in 2007.

The results were presented April 11 as an expedited publication in Neurology, the April edition of the journal Muscle and Nerve from the American Association of Neuromuscular and Electrodiagnostic Medicine, as well as the April issue of PM&R, the American Academy of Physical Medicine and Rehabilitation's scientific journal.

About 16 percent of some 25 million people with diabetes in the United States suffer from PDN, yet it is often unreported and untreated, with an estimated two out of five patients not receiving any care, according to the report.

Lead author Vera Bril, MD, professor of neurology at Toronto General Hospital in Ontario, Canada, discussed with Neurology Today the findings and what they mean for patients and their physicians

HOW DID THE PANEL CATEGORIZE STUDY QUALITY?

First we did a literature search of all relevant studies between 1960 and August 2008, and 2,234 citations were identified. The study titles and abstracts of each were reviewed by at least two panel members for relevance, resulting in 463 articles, each of which was reviewed. In the end 79 relevant studies were identified, and these were reviewed and rated by two members using the AAN criteria for the classification of therapeutic articles. Recommendations were linked to the strength of evidence. If the two reviewers disagreed on a study, a third member evaluated it.

WHAT IS THE MOST SIGNIFICANT FINDING IN TERMS OF TREATMENT?

Pregabalin was the only drug with Level A evidence, based on supportive studies that met all of the criteria for thorough methodology and analysis showing that it can reduce pain in PDN and improve quality of life. But doctors should determine if it is appropriate on a case-by-case basis.

There are a lot of options for patient care. Quite a few of the studies had good Level B data, especially antiepileptics and antidepressants. We found a wide variety of agents with a good evidence base, but it all depends on how each patient responds.

WHAT ARE THE OTHER TREATMENT OPTIONS, AND HOW DID THEY RANK IN TERMS OF POTENTIAL EFFECTIVENESS AND SIDE EFFECTS?

Level B evidence supports the use of gabapentin and sodium valproate, antidepressants such as venlafaxine, duloxetine and amitriptyline, and painkillers such as opioids and capsaicin, although each can cause side effects.

The difference between Level A and Level B status reflects the quality of available studies, based on their criteria and methodology, whether they were multicenter studies using placebo, follow-up, and other issues, with Class I studies being the best. We did not rank the effectiveness of any of the agents, but none of the Level B agents fared any better than another.

In general, efficacy data on each of them were very similar, and physicians could start with any of them. Using opioids, however, can be quite difficult because patient tolerance increases the longer they use them, requiring greater dosages to achieve relief. Also, opioids cause constipation, which is also very common in diabetic patients, so their use can exacerbate the problem, and I tend to avoid them.

There have not been enough one-to-one studies, so we cannot really rank their comparative effectiveness — but it tends to be relatively poor. If patients are not comfortable with pregabalin, we recommend trying the Level B options, but it depends on the individual patient's response.

IN THE RESEARCH, HOW MUCH RELIEF CAN PATIENTS EXPECT FOR TREATMENT?

It depends on the patient. In general, none of the medications results in really good relief, but it can be in the 30 percent to 50 percent range. That means if a patient says the pain is six on a scale of ten, treatment, at best, may reduce the pain to a three.

WHAT ARE THE POTENTIAL SIDE EFFECTS OF PREGABALIN?

Fluid retention and weight gain are the most common problems, but patients can suffer peripheral edema as well. More than 10 percent experience drowsiness or dizziness, while 1 percent to 10 percent have visual disturbances, ataxia, speech problems, tremor, lethargy, memory problems, and constipation.

Less common are reports of depression, confusion, agitation, hallucinations, and tachycardia.

A 2009 AAN GUIDELINE FOUND TRANSCUTANEOUS ELECTRICAL NERVE STIMULATION (TENS) OFFERED NO BENEFIT FOR THESE PATIENTS, BUT THE NEW GUIDELINE DID. WHY THE CHANGE?

The 2009 AAN guideline gave a Level B recommendation for the use of TENS in painful diabetic neuropathy, similar to our guideline. The 2009 guideline failed to find evidence for the use of TENS in chronic low back pain, a different condition than painful diabetic neuropathy.

ARE THERE DATA THAT ANY OF THE OTHER NONPHARMACOLOGIC OPTIONS ARE EFFECTIVE?

We found no evidence that any of these others, such as magnetic shoe inserts, low-intensity laser treatment, Reiki massage, exercise, or acupuncture, had any evidence for efficacy in treating patients.

IS THERE ANY EVIDENCE THAT LONG-TERM TREATMENT WITH PREGABALIN OR ANY OF THE OTHER THERAPIES BECOME LESS EFFECTIVE OVER TIME?

We are unable to answer that question at this time because all of these studies tend to be short-term trials. We know that there is escalation in opioid patients due to increased tolerance, but this is not confirmed with any if the others.

WHAT TYPES OF NEW RESEARCH WOULD BE MOST HELPFUL?

It would help if we had standardized measures of pain, and longer studies that include quality of life and physician function evaluations. PDN needs to be recognized as being a large and common problem, not only in the US, but also throughout the world.

Developing official AAN guidelines will help by giving physicians an evidence-based framework within which to work.

One study showed that while 16 percent of patients suffer chronic neuropathic pain, 12 percent said they had never told their physician, and 39 percent had never been treated for pain.

WHAT DID THE PANEL CONCLUDE ABOUT LONG-TERM RESEARCH NEEDS?

As I said previously, a formal process for rating pain scales is needed for use in all clinical trials, and trials should be expanded to include effects on quality of life and physical function when evaluating the efficacy of new interventions. Further, these measures should be standardized. Future trials should also include head-to-head comparisons of different medications and combinations, and because PDN is chronic, longer trials are needed. Also, standardized metrics for measuring side effects to qualify the effects of interventions need to be developed, and cost-effectiveness studies of different treatments should be performed.

Finally, the actual mechanism of action of electrical stimulation remains unknown, and a better understanding of its role, application, and other aspects of its use should be studied.

Editor's Note: A Recommendation of Caution on the New Guidelines

Neurology Today Associate Editor Robert Holloway, MD: The new guideline on the treatment of painful diabetic neuropathy provides a Level A recommendation for pregabalin and Level B recommendations for 11 drugs, including gabapentin, duloxetine, venlafaxine, amitriptyline, sodium valproate, opiates, and dextromethorphan, among others. Based on the mapping of the evidence available to the recommendation, pregabalin is afforded a “should be offered” treatment recommendation (Level A), while the others received a “should be considered” (Level B).

I believe that the Level A recommendation for pregabalin compared to the Level B recommendations for the others may reflect more about the guideline development process than actual efficacy and safety differences across the treatments.

The guideline development process uses explicit rules to map recommendations from the quantity and quality of evidence available See Levels of Evidence Table. Based on evidence criteria used by the AAN, a Level A recommendation results when there are at least two or more consistent Class I studies. Pregabalin had 3 Class I and 1 Class II studies compared to at least one Class I study or two consistent Class II studies for a Level B recommendation; duloxetine and amitriptyline had 1 Class I and 2 Class II studies; gabapentin and venlafaxine had 1 Class I and 1 Class II studies. Is the weight of one additional Class I study sufficient to change a recommendation from “could” to “should”? This question is important given the use of the pre-specified criteria for distinguishing Class I and II studies. For example, a large randomized clinical trial with 79 percent of the participants completing the study would be considered a Class II study, whereas the exact same trial with a completion rate of 80 percent becomes a Class I study.

Are we demanding too much from the evidence? The heterogeneity and the many nuances involving chronic pain clinical trials make study to study comparisons difficult. In fact, the authors point out that the placebo effect varied from 0 percent to 50 percent for pain reduction across the studies despite generally small to moderate incremental reductions in pain and other patient-reported outcomes for most of the therapies evaluated.

Moreover, there was limited attention to safety and tolerability in formulating these guidelines. To recommend, for example, amitriptyline, dextromethorphan, duloxetine, and sodium valproate on an equal footing is discordant with existing treatment guidelines for neuropathic pain and may place patients at an increased risk of serious adverse events.

Therefore, I recommend that the “should be offered” versus “should be considered” distinction be de-emphasized when looking across the attributes of these agents. It is important for practitioners using this guideline to appreciate that distinctions between Level A and B evidence are but one component of decision making regarding appropriate initial therapy for neuropathic pain in diabetic peripheral neuropathy. Other important factors include safety, cost, patient preference, comorbidities, and physician comfort.

http://journals.lww.com/neurotodayonline/Fulltext/2011/05050/New_Evidence_Based_Guidelines_for_Painful_Diabetic.1.aspx