Am Fam Physician. 2008;78(3):316-320
Original Article: Bell's Palsy: Diagnosis and Management
Issue: October 1, 2007
Available at: https://www.aafp.org/afp/20071001/997.html
to the editor: The article on Bell's palsy provided a good review of the diagnosis and management of this condition. We would like to add a note about the treatment of Bell's palsy with antivirals. The authors stated that although a 2004 Cochrane review1 found insufficient evidence to support the use of antivirals alone, two placebo-controlled studies supported the use of valacyclovir (Valtrex)2 or acyclovir (Zovirax)3 in combination with prednisolone for the treatment of Bell's palsy.
These findings conflict with more recent studies4,5 that demonstrate prednisolone to be beneficial, but that do not support a benefit from antivirals. Consequently, I would recommend that in the Strength-of-Recommendation Taxonomy (SORT) table, the SORT ratings for the use of steroids and the use of antivirals to treat Bell's palsy no longer be combined. Although the evidence for the benefit of steroids merits a B rating (inconsistent or limited-quality patient-oriented evidence), evidence regarding antivirals is mixed and conflicting.
to the editor: The article on Bell's palsy by Drs. Tiemstra and Khatkhate was well done, and I would like to add two therapies that are safe and potentially more efficacious than the treatments discussed in the article: prednisone and acyclovir (Zovirax). Injections of B12 have been used for the treatment of Bell's palsy since at least 1959,1 and their efficacy was demonstrated in a 1995 study.2 In this study, 500 mcg of methylcobalamin was given intramuscularly three times weekly until symptoms improved, or eight weeks. The time to recovery in those receiving methylcobalamin averaged approximately two weeks, while the recovery time in those receiving prednisone averaged more than nine weeks (P < .001). Considering the low cost, lack of side effects, and huge potential for benefit, there seems little reason to refuse treatment with B12 injections for patients with Bell's palsy. The methylcobalamin used in the 1995 study2 may be difficult to find other than having it made by compounding pharmacies, though conventional B12 injections are certainly worth a shot.
Research also shows benefit with using hyperbaric oxygen therapy, which reduced the recovery time from 34 days with steroid treatment to 22 days with use of hyperbaric oxygen alone.3
in reply: Dr. Irani correctly notes that the use of antiviral therapy remains controversial for the treatment of Bell's palsy. Unfortunately, the definitive answer has not yet been provided. Two studies by Hato and colleagues (a 2003 study of acyclovir1 and a 2007 study of valacyclovir2) found that a greater effect of antiviral therapy was demonstrated in more severe cases. The study by Sullivan and colleagues3 that found no effect of acyclovir had less severe cases on average than the conflicting study by Hato and colleagues,1 and did not separately analyze patients with initial complete paralysis. Furthermore, both of the acyclovir studies used the dose recommended for herpes simplex treatment (2,000 mg per day), which is one half the dose recommended for herpes zoster. Because varicella zoster virus has been implicated as one of several potential etiologies, it may be that a higher dose is needed, or that a prodrug such as valacyclovir is needed that provides greater bioavailability of active acyclovir. Results may be affected by the relative incidence of various viral etiologies in different populations. The 2007 article by Kawaguchi and colleagues is interesting. Although they reported no overall significant response to valacyclovir/prednisolone compared with prednisolone alone, they found that patients treated with the combination had a recovery rate 2.5 times higher than the patients receiving prednisolone only if they had serologic evidence of herpes simplex virus reactivation, and 1.6 times higher if they had serologic evidence of varicella zoster reactivation, although the numbers were too small to demonstrate statistical significance.4
Therefore, I believe the weight of evidence for the use of valacyclovir or acyclovir in combination with a corticosteroid still merits a Strength-of-Recommendation Taxonomy (SORT) level B rating for inconsistent and limited patient-oriented evidence. Further, I recommend and routinely treat patients with valacyclovir or acyclovir and prednisone. Permanent facial paresis can result in significant patient distress as well as ophthalmologic complications. One-week treatment with these antivirals is safe and relatively low-cost, and there is no way to initially identify the subset of patients with herpes simplex virus reactivation who might more likely benefit from this treatment. Given the results of the Sullivan study3, it may be reasonable to assign prednisolone therapy a SORT level A rating.
Dr. Sickels cites the only study available regarding methylcobalamin treatment of Bell's palsy.5 However, the study has several methodologic flaws. It was not randomized or blinded, and the group that received prednisone had a mean recovery time of 9.6 weeks (well beyond the roughly three weeks reported in many other large trials even with no treatment) that suggests this group contained more severe cases initially. The standard deviations for the control and prednisone arms are well over 50 percent of the mean recovery times reported, suggesting that their study was poorly powered. This is not surprising given the small number of patients (60 patients divided among three arms). There was also no documentation of whether these patients were vitamin B12-deficient at baseline. Because this treatment has never been studied further, it cannot be recommended.
My main concern with the study Dr. Sickels cites on hyperbaric oxygen is that, although the prednisone group was smaller than the hyperbaric oxygen group (37 versus 42 patients, respectively), it had almost twice as many patients with initial abnormal results on the altered nerve excitability test (nine versus five, respectively) and thus had more patients with more severe disease.6 This suggests that the researchers did not achieve comparable groups through whatever randomization process was used. This treatment also has not been studied further, and thus cannot be recommended.