J Neurol (2008) 255:1726–1730

DOI 10.1007/s00415-008-0008-6 ORIGINAL COMMUNICATION

Martina Minnerop Bell’s palsy

Martin Herbst
Rolf Fimmers Combined treatment of famciclovir
Bertfried Matz and prednisone is superior to prednisone alone
Thomas Klockgether
Ullrich Wüllner

R. Fimmers, PhD with prednisone and famciclovir

Received: 18 December 2007
Received in revised form: 29 March 2008
Accepted: 15 April 2008
Published online: 3 September 2008
M. Minnerop, MD (쾷) · T. Klockgether, MD
· U. Wüllner, MD
Dept. of Neurology
University Hospital of Bonn
Sigmund-Freud-Str. 25
53105 Bonn, Germany
Tel.: +49-228/287-16130
Fax.:+49-228/287-11511
E-Mail: m.minnerop@uni-bonn.de
M. Minnerop, MD
Brain Imaging Center West (BICW)
Leo-Brandt-Str. 1
52425 Jülich, Germany
M. Herbst, MD
Max-Delbrück-Center for Molecular
Medicine
Robert-Rössle-Str. 10
13092 Berlin, Germany
Institute of Medical Biometry, Informatics,
and Epidemiology
University of Bonn
Sigmund-Freud-Str. 25
53105 Bonn, Germany
B. Matz, MD
Institute of Virology
University of Bonn
Sigmund-Freud-Str. 25
53105 Bonn, Germany
■ Abstract There is insufficient
evidence concerning the efficacy of
antiviral treatment of Bell’s palsy
(BP). We therefore compared the
efficacy of prednisone and famci-
clovir to prednisone treatment
alone in BP. A total of 167 consecu-
tive patients with untreated acute
BP were included. Severity of BP
was evaluated using the House-
Brackmann scale (HBS) and virus
antibody tests (herpes simplex
virus, varicella zoster virus) were
performed. Patients admitted on
even dates were treated with pred-
nisone (“P group”) and patients
admitted on odd dates were treated
(“P+F group”). 117 patients com-
pleted the follow-up after 3 months
or later (67 P/51 P+F). While most
patients showed at least partial re-
covery with both treatment types,
improvement of at least 4 grades in
the HBS was more common in the
“P+F group” (29.4 % vs. 11.9 %),
whereas smaller changes of less
than 3 grades were more common
in the “P group” (29.9 % vs. 17.6 %;
Chi-square test, p = 0.02). Patients
with complete BP (HBS grade of 5
or 6) had significantly better
chances of reaching normal func-
tion if treated with famciclovir
additionally instead with pred-
nisone alone (73.7 % vs. 47.1 %;
Cochran-Armitage trend test,
p = 0.03). These results suggest that
the combined treatment of famci-
clovir and prednisolone should be
considered (at least) in patients
with severe BP.
■ Key words Bell’s palsy ·
prednisone · famciclovir · antiviral
therapy

BP is presently unknown. Some studies point to herpes

Introduction family viruses as causative agents: the HSV-1 genome
was found in the geniculate ganglion post mortem in hu-
Bell’s palsy (BP) is the most common cranial nerve le- mans and in endoneural fluid from the facial nerve in
sion, occurring with an annual incidence of 20:100 000 patients with Bell’s palsy [14, 20]. These results support
JON 3008

[4]. Although the majority of patients recover, up to 16 % the earlier proposed hypothesis of an HSV-1 reactiva-
of patients retain persistent sequelae [15]. The cause of tion within the geniculate ganglion with subsequent in-

flammation and entrapment of the nerve at the meatal
foramen or in the labyrinthine segment as pathogenic
mechanism of BP [13].
The antiviral drug aciclovir or its analogue, valaciclo-
vir, has been applied in various trials with inconsistent
results and a Cochrane review in 2004 concluded that
treatment efficiency could not be assessed with the pres-
ent data [2]. Interestingly, the same holds true for the use
of steroids although most national and international
guidelines recommend steroids for management of BP
[9, 17]. Famciclovir, the most recent aciclovir analogue,
offers important advantages in virostatic therapy with
excellent oral bioavailability and longer intracellular
half-life of its active metabolite than aciclovir in VZV-
infected cells. [16, 18].
Therefore, we performed a controlled and prospec-
tive trial comparing the therapeutic effect of prednisone
and famciclovir versus prednisone alone in BP. A pla-
cebo group was not included in our study as placebo
treatment of BP would not comply with current guide-
lines. Clinical outcome and virus antibody titer were
compared between both study arms.
Patients and methods
Consecutive adult patients (age > 18 years) with an untreated acute
facial palsy (onset < 5 days), admitted to our hospital between January
2001 and June 2005, underwent complete neurological and otological
examination. Laboratory testing including cerebrospinal fluid (CSF)
and serum examination were performed. CSF and serum samples
were analyzed for antibodies (ab’s) against Borrelia burgdorferi and
various viruses, e.g., herpes simplex virus (HSV), varicella zoster vi-
rus (VZV), cytomegalovirus, Epstein-Barr virus, tick-borne encepha-
litis, measles, mumps, rubella and enterovirus using standard com-
plement fixation (CFA) and enzyme-linked immunosorbent assays.
Patients with Lyme disease, zoster oticus, or other symptomatic
causes were excluded. The severity of BP was evaluated using the
House-Brackmann scale (HBS) [11]. Therapy was started within 5
days after first symptoms had been noticed. Contraindications to ste-
roid or famciclovir therapy were peptic ulcer, insulin-dependent dia-
betes mellitus, renal or hepatic dysfunction, immune suppression,
and pregnancy.
Patients admitted on even dates were assigned to prednisone (“P
group”), patients admitted on odd dates to prednisone and famciclo-
vir (“P+F group”); treatment was initiated by the resident on call.
Prednisone was administered orally once a day (1 mg/kg body weight)
for 4 days and tapered subsequently over the following 8 days; famci-
clovir was administered orally three times a day (250 mg) for 7 days.
All patients received H2-receptor blockers, artificial tears and oph-
thalmic ointment for eye care.
A follow-up visit was scheduled 3 months after the onset of BP. All
patients were clinically re-examined by a single investigator (MM,
blind to the initial treatment) and serum samples were drawn for pos-
sible virus ab-titer changes.
The study was approved by the ethics committee of the University
Clinics Bonn, and informed consent was obtained from all patients.
Differences in HBS between treatment groups were compared us-
ing the Cochran-Armitage trend test. The Mann-Whitney test or t-
test was used to compare quantitative variables between the treat-
ment groups. Spearman’s correlation coefficient was used to analyze
the correlation between quantitative variables.
1727
Results
A total of 167 patients with BP were included and allo-
cated to treatment according to the day of admission (95
P/72 P+F). Forty-eight patients (30 %), 28 of the “P
group” (29.5 %, initial HBS 3.8 ± 1.1) and 20 of the ”P+F-
group” (27.9 %, initial HBS 3.8 ± 1.0), were lost to follow-
up. Two of them (P+F) discontinued the study because
of adverse effects. A final outcome was available in 117
patients, 67 belonging to ”P group“ and 50 to “P+F
group”.
There were no statistical differences between groups
regarding age, sex or affected side (see Table 1 for patient
characteristics). The initial severity of facial palsy (HBS
grading) was moderate to severe and did not differ be-
tween groups. However, a tendency towards more severe
involvement in the “P+F group” was detected (trend test,
p = 0.12). Possible risk factors for an unfavorable out-
come (diabetes mellitus, arterial hypertension, initial
taste disturbance or post-auricular pain) did neither
correlate with the initial HBS nor differed significantly
between both groups. Initial HBS at first visit did not
correlate with age (Spearman correlation, p = 1.00). El-
derly patients (> 60 years) were equally distributed be-
tween both treatment groups and did not have more
severe palsies than younger patients. No seasonal varia-
tion of BP during the year was observed.
Half of the patients (53 %) attended the scheduled
follow-up appointment; the remaining were seen within
12 months after onset of BP. The interval between fol-
low-up visits did not differ between the two treatment
groups (Table 2, chi-square test: p = 0.81).
At follow-up, 91 % patients in the “P group” and 96 %
in the “P+F-group” had at least some improvement of
voluntary movement. An improvement of at least 4
grades at the HBS was more common in “P+F group”
(28.0 %) than in “P group” (12 %), whereas no or only
minor improvements in the HBS were more common in
Table 1 Baseline characteristics of patients
P group P + F group
N 67 50
Age (years; mean ± SD) 40.6 ± 20.5 42.6 ± 30.9
Older patients (> 60 years) 14 13
Sex (m/f) 34/33 26/24
Affected face side (right/left) 38/29 25/25
Diabetes mellitus (non-insulin dependent) 3 1
Arterial hypertension 9 9
Taste disturbance 19 19
Post-auricular pain 16 11
Initial HBS (mean ± SD) 3.9 ± 1.0 4.1 ± 1.0
Time between onset of symptoms and start of 2.4 ± 1.9 2.6 ± 1.9
treatment (days; mean ± SD)
1728

Table 2 Time between first visit and follow-up visit of 17 patients (“P group”); trend test, p = 0.04; Fig. 2).
3 months 6 months 7–12 months > 12 months Total Older patients in the “P+F group” did not benefit more
from additional famciclovir than younger patients
P Group 36 16 9 6 67 (trend test, p = 0.37). The time between onset of symp-
P+F Group 26 14 8 2 50 toms and start of treatment did not differ between
Total 62 30 17 8 117 groups or correlate with clinical improvement at follow-
up.

the “P group” (30 % compared to 18 % in “P+F group”

(trend-test, p = 0.028; Fig. 1). Patients with severe BP ■ Serological findings
(initial HBS grade of 5 or 6) had a better chance to
achieve normal function if they additionally received VZV-ab titer (CFA) at first visit were available in 40 pa-
famciclovir (72 % of 18 patients (“P+F group”) vs. 47 % tients allocated to the “P group” (60 %) and 35 patients

Fig. 1 Clinical improvement in both

groups according to a reduction in HBS

30%

20%

10%

0%
0 −1 −2 −3 −4 −5
P group 9.0% 20.9% 26.9% 31.3% 10.4% 1.5%
P+F group 4.0% 14.0% 24.0% 30.0% 22.0% 6.0%

Fig. 2 HBS grade at follow-up visit in 80%

both groups of patients with an initial
HBS grade of 5 or 6

60%

40%

20%

0%
1 2 3 4 5
P group 47.1% 17.7% 17.7% 5.9% 11.8%
P+F group 72.2% 16.7% 11.1% 0.0% 0.0%

allocated to the “P+F group” (70 %). Ab-titer change of
VZV and HSV between the first and second visits did
not differ between groups, nor did they correlate with
HBS improvement between visits (Spearman correla-
tion, p = 0.38 (HSV), p = 0.63 (VZV)).
■ Adverse effects
Two patients developed a skin rash, most likely due to
famciclovir and discontinued the medication. These pa-
tients were not included in our statistical analysis. One
young male patient reported subacute onset of hairloss
after completion of the medical treatment.
Discussion
We found a significantly better chance for BP patients to
reach normal function if they were treated with predni-
sone and famciclovir combined instead with prednisone
alone. In particular, patients with severe BP had a more
favorable outcome with additional antiviral therapy.
Some methodological issues however need to be consid-
ered: The use of pseudo-randomization may have led to
bias in our study results. However, comparison of puta-
tive influencing factors detected no differences between
the two groups. The drop-out rate was higher than in
similar studies. This is in part due to the German health
care system where patients find medical advice readily
and may be more reluctant to return to the university
clinic for follow-up visits. The good prognosis of BP in
general may further reduce the motivation for a follow-
up visit, but at least the initial HBS did not differ be-
tween patients with and without follow-up. Further-
more, drop-out rates did not differ between treatment
groups. As we compared the therapeutic benefit between
groups, the high drop-out rate should not have influ-
enced our study results. The different time interval until
follow-up, not differing between groups, may have led to
reduced power but not to a bias preferring one thera-
peutic regime.
With respect to antiviral therapy, available studies
gave contradictory results: compared with prednisone
alone, a combined therapy of aciclovir and prednisone
showed favorable outcome in patients receiving both
[1], whereas a study comparing single aciclovir with
prednisone only therapy resulted in a better clinical out-
come in the prednisone group [5]. A comparison of
prednisone and valaciclovir with a placebo group found
a better clinical outcome only in elderly patients [3],
thought to have a poorer prognosis [15]. A recent large
Japanese study described the combination of valaciclo-
vir and prednisolone superior to prednisolone and pla-
cebo [10], whereas another study using valaciclovir did
not find differences between the combined therapy and
1729
prednisolone alone [12]. Most recently, a British study
found a favorable outcome only in patients treated with
prednisone alone, whereas additional or single antiviral
therapy had no significant effect [19].
The different results in these studies can partly be
attributed to different study designs, but even studies
with a similar design come to different conclusions [10,
12]. The initial severity of BP may further influence
study results [7]. Study patients of Hato et al. [10], show-
ing a positive effect of a combined therapy of predniso-
lone and valaciclovir, appear to have had more severe BP
than those in the study by Sullivan et al. [19], who could
not find a benefit in patients treated with aciclovir alone
or in combination with prednisolone. In our study, the
severity of BP was similar to the study of Hato et al. and
like them we found the benefit of the combined treat-
ment to be dependent on the severity of facial palsy.
The used antiviral substances, dosage, start and dura-
tion of therapy may cause further heterogeneity between
the conducted studies. Aciclovir has a low oral bioavail-
ability of 10–20 % [6] and must be administered five
times daily, which was not complied with in one study
[5]. The systemic availability of aciclovir, already low in
the fasted state, is further compromised if taken with
food. Even if study patients are well instructed, the cor-
rect administration is difficult to monitor. Valaciclovir,
used in both Japanese studies, is a prodrug of aciclovir
and has a higher bioavailability (54 %) compared with
orally administered aciclovir [8]. Famciclovir, the most
recent aciclovir analogue, exhibits excellent oral bio-
availability (60–75 %), not affected by concurrent food
intake. Its active metabolite penciclovir triphosphate
has a much longer intracellular half-life in VZV-infected
cells than aciclovir. Therefore, famciclovir offers impor-
tant advantages in virostatic therapy [16, 18]. In our
study, famciclovir was used for the first time in the treat-
ment of BP.
In fact, a considerable number of our patients bene-
fited from additional antiviral therapy with famciclovir,
yielding circumstantial evidence for a pathogenic rele-
vance of herpesviruses in BP (HSV, VZV or others) at
least in a subset of BP cases.
Beside methodological problems to detect reactiva-
tion of a widespread virus, it is possible that another, yet
unidentified (herpes) virus, might be the causative
agent.
In summary our data suggest that a combined treat-
ment of famciclovir and prednisolone should be consid-
ered (at least) in patients with severe BP. Further larger
double-blind multi-center studies will be essential to
validate the therapeutic benefit of combined treatment
with famciclovir and prednisolone and may help to elu-
cidate a potentially pathogenic role of herpes virus in
BP.
■ Conflict of interest The authors declare no conflict of interest.
1730
References
1. Adour KK, Ruboyianes JM, Von
Doersten PG, Byl FM, Trent CS,
Quesenberry CP Jr, Hitchcock T (1996)
Bell’s palsy treatment with acyclovir
and prednisone compared with pred-
nisone alone: a double-blind, random-
ized, controlled trial. Ann Otol Rhinol
Laryngol 105:371–378
2. Allen D, Dunn L (2004) Aciclovir or
valaciclovir for Bell’s palsy (idiopathic
facial paralysis). Cochrane Database
Syst Rev, Issue 2:CD001869
3. Axelsson S, Lindberg S, Stjernquist-
Desatnik A (2003) Outcome of treat-
ment with valacyclovir and prednisone
in patients with Bell’s palsy. Ann Otol
Rhinol Laryngol 112:197–201
4. Brackmann DE (1974) Bell’s palsy:
incidence, etiology, and results of
medical treatment. Otolaryngol Clin
North Am 7:357–368
5. De Diego JI, Prim MP, De Sarria MJ,
Madero R, Gavilan J (1998) Idiopathic
facial paralysis: a randomized, pro-
spective, and controlled study using
single-dose prednisone versus acyclo-
vir three times daily. Laryngoscope
108:573–575
6. De Miranda P, Blum MR (1983) Phar-
macokinetics of acyclovir after intra-
venous and oral administration. J Anti-
microb Chemother12(Suppl B):29–37
7. Gilden DH, Tyler KL (2007) Bell’s palsy
– is glucocorticoid treatment enough?
N Engl J Med 357:1653–1655
8. Granero GE, Amidon GL (2006) Stabil-
ity of valacyclovir: implications for its
oral bioavailability. Int J Pharm 317:
14–18
9. Grogan PM, Gronseth GS (2001) Prac-
tice parameter: Steroids, acyclovir, and
surgery for Bell’s palsy (an evidence-
based review): report of the Quality
Standards Subcommittee of the Ameri-
can Academy of Neurology. Neurology
56:830–836
10. Hato N, Yamada H, Kohno H, Matsu-
moto S, Honda N, Gyo K, Fukuda S,
Furuta Y, Ohtani F, Aizawa H, Aoyagi
M, Inamura H, Nakashima T, Nakata S,
Murakami S, Kiguchi J, Yamano K,
Takeda T, Hamada M, Yamakawa K
(2007) Valacyclovir and Prednisolone
Treatment for Bell’s Palsy: A Multi-
center, Randomized, Placebo-Con-
trolled Study. Otol Otol Neurotol 28:
408–413
11. House JW, Brackmann DE (1985)
Facial nerve grading system. Otolaryn-
gol Head Neck Surg 93:146–147
12. Kawaguchi K, Inamura H, Abe Y, Koshu
H, Takashita E, Muraki Y, Matsuzaki Y,
Nishimura H, Ishikawa H, Fukao A,
Hongo S, Aoyagi M (2007) Reactivation
of herpes simplex virus type 1 and
varicella-zoster virus and therapeutic
effects of combination therapy with
prednisolone and valacyclovir in pa-
tients with Bell’s palsy. Laryngoscope
117:147–156
13. McCormick DP (1972) Herpes-simplex
virus as a cause of Bell’s palsy. Lancet
29:937–939
14. Murakami S, Mizobuchi M, Nakashiro
Y, Doi T, Hato N, Yanagihara N (1996)
Bell Palsy and Herpes Simplex Virus:
Identification of Viral DNA in Endo-
neurial Fluid and Muscle. Ann Intern
Med 124:27–30
15. Peitersen E (2002) Bell’s palsy: the
spontaneous course of 2,500 periph-
eral facial nerve palsies of different
etiologies. Acta Otolaryngol Suppl
549:4–30
16. Pue M, Benet LZ (1993) Pharmacoki-
netics of famciclovir in man. Antiviral
Chem Chemother 4(Suppl 1):47–55
17. Salinas RA, Alvarez G, Ferreira J (2004)
Corticosteroids for Bell’s palsy (idio-
pathic facial paralysis). Cochrane
Database Syst Rev, Issue 2:CD001942
18. Standring-Cox R, Bacon TH, Howard
B, Gilbart J, Boyd MR (1994) Prolonged
activity of penciclovir in cell culture
against varicella-zoster virus. Antiviral
Res 23(Suppl 1):96
19. Sullivan FM, Swan IR, Donnan PT,
Morrison JM, Smith BH, McKinstry B,
Davenport RJ, Vale LD, Clarkson JE,
Hammersley V, Hayavi S, McAteer A,
Stewart K, Daly F (2007) Early treat-
ment with prednisolone or acyclovir
in Bell’s palsy. N Engl J Med 357:
1598–1607
20. Takasu T, Furuta Y, Sato KC, Fukuda S,
Inuyama Y, Nagashima K (1992) Detec-
tion of latent herpes simplex virus
DNA and RNA in human geniculate
ganglia by the polymerase chain reac-
tion. Acta Otolaryngol 112:1004–1011