Medical Treatment and Grading of Bell's Palsy

Full text

(1)

(2)

(3)

(4)

(5)

(6) .

(7) AND'RADINGOF !".

(8)

(9) . # $%&%' # $()*(%&&)&%+ ,-.- -,,- / %().

(10) .

(11)

(12)

(13)

(14) !" #$ %& % ' ( &))**! &+,&%-

(15) $

(16) -

(17)

(18)

(19) -.$

(20)

(21) $/'

(22) -0 123$

(23) 4

(24) 4$2 5 32)**!20 3 6

(25) -57. 2" . 2

(26)

(27)

(28) 89*282 2! !&%%8%8&+2 3$

(29) - $ $ 4

(30) $ --

(31) -

(32)

(33)

(34)

(35) -57 23$

(36) -

(37) $ . $-

(38) $ $- 4

(39)

(40) $

(41) -

(42)

(43) 4

(44) 2:

(45) $

(46) 4 $

(47)

(48) - $

(49) -- $

(50) -

(51) 2' $

(52) $ 4 $

(53)

(54) ;

(55) 5 < $ - 4 2 '

(56) 0 )**&

(57) )**

(58)

(59)

(60)

(61)

(62)

(63) 4 $ &)

(64) $ -

(65)

(66) 4 4 -

(67) 4 $ 57 2 =- +!

(68) )! 4 $

(69)

(70) > )*9

(71)

(72) )&*

(73)

(74)

(75) )*

(76)

(77) )*9

(78)

(79)

(80) 23

(81)

(82) 4 - $

(83) $8&9 4$

(84)

(85)

(86)

(87) 4 $ $8&+4$

(88)

(89) /*2***&12 " &)

(90) $+**

(91) -8&9 /)@1 $

(92)

(93)

(94) $

(95)

(96) 4 $)+

(97) -8&+ /%@1 $

(98)

(99)

(100)

(101) /*2***&12 A

(102) 4

(103) -

(104) .

(105) --

(106) -

(107)

(108)

(109)

(110) &)

(111) $2.

(112)

(113) B

(114)

(115)

(116) -- $

(117)

(118) - 2.

(119) - &&

(120) & 4

(121)

(122)

(123) -

(124) -

(125) &)

(126) $2:

(127) -

(128) $

(129) 57 . -- $$

(130)

(131) - $

(132) -

(133)

(134) --

(135) 2 :4$ C . &**

(136)

(137) - 4 $-

(138) . $ 4 $

(139)

(140) ;

(141) 5 < $ 23$$$. 4 -

(142) 4 $

(143)

(144)

(145) < $ 2" -- 4 $

(146)

(147) ;

(148) 5 4

(149) 2 57 -

(150)

(151)

(152)

(153)

(154) ;

(155) 5 < $

(156)

(157) -

(158) 4

(159) !Department of Surgical Sciences, # $%& !

(160)

(161)

(162) ! D3$

(163) 5 )**! E&9%&9)*9 5E! !&%%8%8&+ , ,,, &**!8/$. ,BB 22B

(164) F G , ,,, &**!81.

(165) List of Papers. This thesis is based on the following papers, which are referred to in the text by their Roman numerals. I. Berg T, Jonsson L, Engström M. Agreement between the Sunnybrook, House-Brackmann, and Yanagihara facial nerve grading systems in Bell's palsy. Otol Neurotol 2004;25(6):1020-6.. II. Engström M, Berg T, Stjernquist-Desatnik A, Axelsson S, Pitkäranta A, Hultcrantz M, Kanerva M, Hanner P, Jonsson L. Prednisolone and valaciclovir in Bell's palsy: a randomised, double-blind, placebo-controlled, multicentre trial. Lancet Neurol 2008;7(11):993-1000.. III. Berg T, Axelsson S, Engström M, Stjernquist-Desatnik A, Pitkäranta A, Kanerva M, Jonsson L. The course of pain in Bell’s palsy: treatment with prednisolone and valacyclovir. Accepted for publication 2009 in Otol Neurotol.. IV. Berg T, Marsk E, Engström M, Hultcrantz M, Hadziosmanovic N, Jonsson L. The effect of study design and analysis methods on recovery rates in Bell’s palsy. Submitted for publication.. Reprints were made with permission from the respective publishers..

(166)

(167) Contents. Introduction.....................................................................................................7 Anatomy of the facial nerve .......................................................................7 Acute peripheral facial palsy......................................................................8 History ...................................................................................................8 Etiological factors..................................................................................8 Grading facial function ........................................................................10 Bell’s palsy...............................................................................................13 Etiology ...............................................................................................13 Treatment.............................................................................................15 Pain in Bell’s palsy ..............................................................................18 Recovery rates in Bell’s palsy .............................................................18 Aims of the study ..........................................................................................19 Material and Methods ...................................................................................20 Patients .....................................................................................................20 Methods....................................................................................................21 Statistical analysis ....................................................................................22 Results...........................................................................................................24 Paper I ......................................................................................................24 Paper II .....................................................................................................25 Paper III....................................................................................................29 Paper IV ...................................................................................................32 Discussion .....................................................................................................35 Conclusions...................................................................................................39 Acknowledgements.......................................................................................40 References.....................................................................................................43.

(168) Abbreviations. CI DNA HR IQR ITT MRI HSV PCR VAS VZV. Confidence interval Deoxyribonucleic acid Hazard ratio Interquartile range Intention to treat Magnetic resonance imaging Herpes simplex virus Polymerase chain reaction Visual analogue scale Varicella zoster virus.

(169) Introduction. Anatomy of the facial nerve A general knowledge of the anatomy of the seventh cranial nerve is essential for diagnosis and treatment of facial nerve disorders (May and Schaitkin 2000). The greater part of the facial nerve is composed of motor fibres to the facial muscles. In addition, the nerve carries secretomotor fibres for the submandibular, sublingual, and lacrimal glands. The facial nerve also has two sensory, or afferent, components. One carries taste sensation from the anterior two thirds of the tongue and the palate, while the other transmits ordinary sensation from the skin in the region of the external ear (Diamond and Frew 1979, May and Schaitkin 2000). The motor face area of the cerebral cortex is situated on the pre- and postcentral gyri. Discharges from this area are carried through fascicles of the corticobulbar tract to the lower brainstem where they synapse in the facial nerve nucleus located in the pons. The corticobulbar tracts arising from the upper face area cross and recross in reaching the facial motor nucleus, while the tracts to the lower face only cross. This anatomical arrangement allows the clinical distinction between a central and peripheral nerve dysfunction (Crosby and Dejonge 1963, Malone and Maisel 1988, May and Schaitkin 2000). The facial nerve emerges from the brainstem at the pontomedullary junction and enters the internal auditory meatus. It travels further into the bony fallopian canal, which is subdivided into three segments based upon changes of direction: the labyrinthine, tympanic, and mastoid segments. The labyrinthine segment lies between the vestibule and cochlea and contains the geniculate ganglion. At the geniculate ganglion, the greater and lesser petrosal nerves arise. The remaining fibres continue to the tympanic and mastoid segments before the nerve descends to the stylomastoid foramen. In the mastoid segment, the facial nerve has three branches: the nerve to the stapedius muscle, the chorda tympani nerve, and the auricular branch of the vagus nerve (Malone and Maisel 1988, May and Schaitkin 2000). As the nerve exits the stylomastoid foramen, the posterior auricular nerve forms the first branch. The facial nerve then passes forward to the parotid gland where it bifurcates into an upper and lower division. Within the parotid gland, the facial nerve subdivides into the five terminal branches innervating the facial muscles: the temporal, zygomatic, buccal, marginal mandibular and cervical branches (Malone and Maisel 1988). 7.

(170) Acute peripheral facial palsy History The first medical publications on facial nerve paralysis appeared in the 18th century. In 1798, Friedrich published his observations on three patients with facial nerve paralysis (Bird 1979). Charles Bell presented his discovery of the seventh cranial nerve and its role in innervation of the facial muscles in 1821 (Bell 1821). To support his thesis, Bell presented a series of cases of facial paralysis that resulted from direct trauma or infection (Bell 1829). His name was subsequently applied to the acute idiopathic peripheral facial palsy, so-called Bell’s palsy. In 1870, Tryde published the first report on facial palsy in association with herpes zoster (Tryde 1870). Later, J. Ramsay Hunt suggested that herpes zoster oticus is a result from a geniculate ganglionitis (Hunt 1907). This syndrome, which now carries his name, is characterized by facial paralysis, severe pain, and ipsilateral vesicles (May and Schaitkin 2000).. Etiological factors Acute peripheral facial palsy is a diagnostic challenge, and often one cannot determine the etiology merely from its onset. The differential diagnostic possibilities are numerous and the most common causes are listed in Table 1. Peitersen studied 2,570 cases with peripheral facial nerve palsy during a period of 25 years. In his material, 349 of the patients (14%) were aged less than 15 years. Peitersen classified 1,701 patients as Bell’s palsy (66%) and 869 patients as non-Bells palsy (34%). Of the 869 non-Bell’s palsy patients, palsies in neonatal age (19%), herpes zoster (13%), trauma (11%), diabetes mellitus (9%), pregnancy (5%), polyneuritis including Borrelia (5%), and parotid tumors (5%) were the most common etiological and/or concomitant factors in the disease (Peitersen 2002). In endemic areas, Borrelia burgdorferi has been reported to be the cause of palsy in 10% of adults (Ljøstad et al. 2005). Furthermore, Tveitnes and co-workers examined 115 children with acute peripheral facial palsy over a 9 year period in another endemic area of Norway, and 75 (65%) of these were diagnosed as Lyme borreliosis based on lumbal puncture (Tveitnes et al. 2007).. 8.

(171) Table 1. Causes of Facial Paralysis (modified from May and Schaitkin 2000). Birth Congenital Acquired Trauma Basal skull fracture Facial injuries Neurologic Multiple sclerosis Guillan-Barré syndrome Infection External otitis Otitis media Mastoiditis Herpes zoster cephalicus Encephalitis Poliomyelitis Mumps Mononucleosis Tuberculosis Lyme disease Parotitis Metabolic Diabetes mellitus Hyper- and hypothyreosis Pregnancy Neoplastic Cholesteatoma Seventh nerve tumour Glomus jugulare tumour Meningioma Benign and malignant parotid lesions Toxic Iatrogenic Surgery Embolisation Idiopathic Bell’s palsy Melkersson-Rosenthal syndrome Amyloidosis Sarcoidosis. 9.

(172) Grading facial function Grading facial function is necessary for evaluating and communicating the spontaneous course of and the results of medical and surgical treatment of facial palsy (House 1983, House and Brackmann 1985, Croxson et al. 1990). There are two main types of facial grading systems; gross and regional systems. Gross clinical 5-6-point scales with an overall assessment of facial motor function have been proposed (Botman and Jongkees 1955, Peitersen 1977, May et al. 1981, House and Brackmann 1985). Regional unweighted and weighted systems, evaluating different areas of facial function, have been devised by other authors (Janssen 1963, Yanagihara 1977, Smith et al. 1992, Lewis and Adour 1995, Ross et al. 1996). In addition to these main systems, specific and/or objective scales also exist (Stennert et al. 1977, Burres and Fisch 1986, Murty et al. 1994, Kang et al. 2002). The House facial nerve grading system was proposed by House in 1983 (House 1983). After minor modifications, the system was presented by House & Brackmann in 1985 (House and Brackmann 1985). The HouseBrackmann system was adopted as a standard of grading facial function by the Facial Nerve Disorders Committee of the American Academy of Otolaryngology-Head and Neck Surgery. This system is based on a 6-grade score, (I to VI) that offers a gross evaluation of facial motor function and also includes evaluation of sequelae (Table 2). It has become the most universally adopted scaling system for facial nerve disorders in the USA and Europe. Table 2. House-Brackmann facial nerve grading system I to VI (summarised).. I II III IV V VI. Normal Mild dysfunction; slight weakness noticeable only on close inspection Moderate dysfunction; obvious, but not disfiguring, difference between the two sides Moderately severe dysfunction; obvious weakness and/or disfiguring asymmetry Only barely perceptive motion Loss of tone. The regional Yanagihara grading system, presented by Yanagihara in 1976 (Yanagihara 1977), assesses 10 separate aspects of function in the different facial muscles (Table 3). Each function is scored from 0 to 4, giving a maximum score of 40. This grading scale does not include any secondary effects. Yanagihara is the most widely used system in Japanese studies for evaluating facial nerve function in Bell’s palsy, herpes zoster oticus and following acoustic neuroma surgery (Satoh et al. 2000).. 10.

(173) Table 3. Yanagihara 5-point facial nerve grading system 0 to 40. The scale consists of normal function, slight paralysis, moderate paralysis, severe paralysis, and total paralysis for which points 4, 3, 2, 1 and 0, respectively, are awarded. At rest Wrinkle forehead Blink Slight closure of eye Tight closure of eye Closure of eye on involved side only Wrinkle nose Whistle Grin Depress lower lip. Scale of five rating 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2 0 1 2. 3 3 3 3 3 3 3 3 3 3. 4 4 4 4 4 4 4 4 4 4. In 1996, Ross et al. (Ross et al. 1996) proposed a grading system designated the Sunnybrook facial grading system. This is a regional weighted system based on evaluation of resting symmetry, degree of voluntary excursion and incorporation of secondary defects (synkinesis) to form a maximum composite score of 100 (Table 4). Its purpose was to develop a clear, well-defined system that provides an accurate description of facial motor function and that is responsive to clinically important changes (Ross et al. 1996). The Sunnybrook system reports facial function in a more continuous manner and has a wider response range than House-Brackmann (Ross et al. 1996). The intrarater and inter-rater reliability of the Sunnybrook system is high, both when applied by novice and expert users (Hu et al. 2001), and is more reliable than House-Brackmann (Kanerva et al. 2006).. 11.

(174) 0 1 1. 0 2 1 1. 0 1 1 1 1. 1 1 1. Gentle eye closure. Open mouth smile. Snarl. Lip packer. Slight movement 2. 2. 2. 2. 2. 3. 3. 3. 3. 3. Mild excursion. 4. 4. 4. 4. 4. Voluntary movement score:. 1. No movement. Forehead wrinkle. 5. 5. 5. 5. 5. Movement complete. 1. 1. 1. 1. 1. 2. 2. 2. 2. 2. Synkinesis score:. 0. 0. 0. 0. 0. Voluntary movement score x 4 – Resting symmetry score x 5 – Synkinesis score x 1 = Composite score. normal narrow wide eyelid surgery Cheek (naso-labial fold) normal absent less pronounced more pronounced Mouth normal corner drooped corner pulled up/out Resting symmetry score:. Eye. Movement almost complete. Synkinesis. None. Symmetry of Voluntary Movement. Mild. Resting Symmetry. Moderate. Table 4. The Sunnybrook facial nerve grading system. This regional weighted system is based on evaluation of different facial regions including resting symmetry, symmetry of voluntary movement and synkinesis to form a composite score from 0 to 100. Summarized from Ross et al. (1996).. 3. 3. 3. 3. 3. Severe.

(175) Bell’s palsy Bell’s palsy is unilateral weakness or paralysis of the face due to acute peripheral facial nerve dysfunction, with no readily identifiable cause, and with some recovery of function within six months (May and Hughes 1987). It is a diagnosis of exclusion and trauma, otitis media, borreliosis, herpes zoster infection and neoplastic lesions of the temporal bone or parotid gland have therefore to be ruled out. Bell’s palsy represents 70% of the cases with acute peripheral facial palsy and the incidence is about 30 per 100,000 inhabitants (Katusic et al. 1986, Yanagihara 1988, Peitersen 2002). Approximately 70% of the patients with Bell’s palsy recover completely within six months without treatment. The remaining 30% suffer sequelae, including residual paresis, contracture, and synkinesis (Peitersen 2002).. Etiology Many theories have been proposed to explain the origin of Bell’s palsy. Ischemia, immunological reactions, and viral infections have all been suggested to be a part of the etiology, for example. Some authors have suggested that ischemia (due to disturbed circulation in the vasa nervorum) leads to the nerve injury in Bell’s palsy. Vascular spasm causes a swelling of the nerve in the Fallopian canal, and secondary compressive edema ensues (Miehlke et al. 1981). This theory is the background for the surgical decompression in the treatment of the disease. An immunological hypothesis was introduced by McGovern and coworkers based on their experimental work on animals (McGovern et al. 1972, McGovern et al. 1977). Later, Hughes et al. demonstrated abnormal lymphocyte transformation in patients with Bell’s palsy and suggested that some instances of Bell's palsy result from cell-mediated immunity against peripheral nerve antigens (Hughes et al. 1986). Their results encouraged further research into steroid and other immunotherapies. During recent decades, the viral theory has gained the most interest, and especially the neurotropic herpes viruses, which are known to become latent after primary infection, are considered important candidates in disease development. In 1972 McCormick (McCormick 1972) suggested herpes simplex virus (HSV) as a cause of Bell’s palsy. His theory was supported by the results in the serological study of HSV antibodies by Adour and co-workers in 1975 (Adour et al. 1975). Even more compelling data linking HSV to Bell’s palsy were presented by Murakami et al. in 1996 (Murakami et al. 1996). Endoneurial fluid from the facial nerve and biopsies from the posterior auricular muscle were tested for HSV-1 DNA and varicella zoster virus (VZV) DNA using polymerase chain reaction (PCR) in 14 Bell’s palsy patients undergoing decompression surgery. HSV-1 genomes were detected in 11 of 14 patients and VZV in none. They concluded that HSV-1 is the major 13.

(176) etiologic agent in Bell’s palsy (Murakami et al. 1996). This finding was further supported by Furuta et al. who found significantly increased levels of HSV DNA in the saliva from 47 patients with Bell’s palsy (Furuta et al. 1998). However, more recent studies have failed to support this theory and the role of HSV in the disease is still unclear (Linder et al. 2005, StjernquistDesatnik et al. 2006). In the study by Stjernquist-Desatnik et al., biopsies were taken from the posterior auricular muscle within 72 hours after the onset of palsy and cerebrospinal fluid was analyzed by nested PCR for HSV1 and VZV DNA in 20 patients. They detected HSV-1 DNA in only one of the muscle biopsies and none of the cerebrospinal fluid samples. Linder et al. (Linder et al. 2005) performed PCR specific for DNA of HSV-1, HSV-2, and VZV in facial muscle biopsies from patients with Bell's palsy. As control specimens, the Scarpa’s ganglion of patients with Meniere's disease and the geniculate ganglion harvested at autopsy from patients without history of facial palsy were used. The investigators were not able to detect HSV or VZV genomic DNA in ganglion scarpae or muscle biopsies in Bell’s palsy patients or controls. HSV-1 DNA was detected in 86% and VZV DNA in 43% of the geniculate ganglions from patients without history of facial palsy. Their findings thus contradict the results of Murakami and co-workers (Murakami et al. 1996). Linder et al. concluded that the sole presence of HSV genomic DNA within the geniculate ganglion does not explain a direct association with Bell’s palsy (Linder et al. 2005). Other studies suggest that VZV plays an important etiological role in Bell’s palsy, causing acute peripheral facial palsy without vesicles, so-called zoster sine herpete. Furuta et al. (Furuta et al. 2000), using the combination of serological assays and PCR, found VZV reactivation in 35 of 121 patients (29%) clinically diagnosed with Bell’s palsy. Based on their studies, they consider VZV to be one of the major etiological agents in Bell’s palsy, both in adults and in children (Furuta et al. 2000, Furuta et al. 2005). On the other hand, after studying the saliva of healthy individuals without facial palsy, Mehta et al. (Mehta et al. 2004) found that VZV can reactivate subclinically in healthy individuals after non-surgical stress. These findings might indicate that the VZV reactivation seen in other studies is a result of the psychological stress caused by Bells’ palsy, and not the cause of the disease. Whatever the cause of the initial damage of the facial nerve in Bell’s palsy, substantial data indicate that the injury causes edema with mechanical compression of the facial nerve within the Fallopian canal. Entrapment of the swollen facial nerve in the narrow meatal segment and in the labyrinthine portion is suggested to be a critical component in the pathogenesis of Bell’s palsy (Fisch and Esslen 1972, Saito 2000, Linder et al. 2005). Based on the MRI findings with gadolinium enhancement of the facial nerve in the meatal and labyrinthine portion, this edema has been suggested to be due to an inflammatory reaction (Schwaber et al. 1990, Tien et al. 1990, Engström et al. 1993, Song et al. 2008). 14.

(177) Treatment Treatment of Bell’s palsy has been a matter of debate for decades and no international consensus has yet been reached. Many different treatment methods have been suggested, but decompression surgery, corticosteroids, and antiviral treatments dominate the literature. However, the high spontaneous recovery rates in Bell’s palsy make evaluation of treatment effect difficult and require a large number of patients to be studied (Grogan and Gronseth 2001, Allen and Dunn 2004, Salinas et al. 2004). Surgical treatment: In 1932, Balance and Duel (Balance and Duel 1932) advocated decompression surgery for Bell’s palsy. Indication for surgery was based on the theory of swelling and entrapment of the facial nerve in the bony Fallopian canal. The number of operations increased dramatically, and several surgical approaches were reported (May 1979, Yanagihara et al. 1979, Fisch 1981). Because of the lack of conclusive trials and complications reported, surgical procedures have declined and decompression surgery is no longer recommended in Bell’s palsy (May and Schaitkin 2000, Grogan and Gronseth 2001, Adour 2002, Peitersen 2002). Corticosteroid treatment: The most commonly prescribed medication for Bell’s palsy has been oral corticosteroids. However, treatment with corticosteroids is controversial and has been a matter of debate for several decades. In a 1954 controlled trial, Taverner (Taverner 1954) treated 14 patients within 10 days after onset of palsy with oral cortisone acetate (200 mg daily for 3 days, 100 mg for 3 days and 50 mg for 2 days) and did not find a significant reduction in the incidence of denervation compared with 12 untreated patients. In a study published in 1972, Adour and co-workers (Adour et al. 1972) compared 194 Bell’s palsy patients treated with prednisone (40 mg daily for 4 days and tapering to 8 mg daily for 8 days) with 110 patients who received no treatment. In patients with a clinically complete paralysis, none in the treated group showed complete denervation (measured by nerve excitability test) compared to 24 percent in the untreated group (p<0.001). In the prospective, controlled, double-blind study conducted between 1972 and 1974 by May et al. (May et al. 1976), 26 patients were treated with prednisone (total of 410 mg over 10 days in descending doses) plus vitamins compared with 25 patients treated with vitamins only. No statistically significant beneficial effect of steroid therapy upon recovery from Bell’s palsy was demonstrated. In the prospective, randomised, controlled study published in 1999 by Ünüvar and colleagues (Ünüvar et al. 1999), 21 children with Bell’s palsy were treated with oral methylprednisolone (1 mg/kg daily for 10 days then gradually withdrawn over 3-5 days) and compared with 21 children in whom no treatment was given. The improvement rates between the treated and untreated children did not differ significantly. In the double-blind, randomised, placebo-controlled study by Lagalla and co-workers in 2002 (Lagalla et al. 2002), 62 consecutive patients were enrolled within 72 hours 15.

(178) of palsy onset and assigned to high-dose intravenous prednisone (1,000 mg daily for three days and then 500 mg daily for three days) in combination with intramuscular polyvitaminic therapy, or polyvitaminic therapy alone. The time needed to recover facial function to House-Brackmann grade III or better was significantly shorter in the 32 treated patients compared with the 30 controls (p=0.005), but this result was not confirmed in the long-term after 6-12 months of follow-up. Reviews of the literature have been made to determine the effectiveness of corticosteroids in Bell’s palsy. The report of the Quality Standards Subcommittee of the American Academy of Neurology (Grogan and Gronseth 2001) noted that the pooled results of the four trials meeting their inclusion criteria (Taverner 1954, May et al. 1976, Brown 1982, Austin et al. 1993) showed significantly better facial outcomes in steroid-treated patients compared with non-steroid-treated patients (relative rate good outcome 1.16, 95%, CI 1.05 to 1.29). It was, however, concluded that a benefit from steroids has not been definitively established, even though the available evidence suggests that they probably are effective (Grogan and Gronseth 2001). In the Cochrane review by Salinas et al. (Salinas et al. 2004), two of studies in the report of Grogan (Grogan and Gronseth 2001) were excluded due to the use of alternation in “matched patients” as the method of randomisation, so-called quasi-randomisation (Brown 1982), and to the exclusion of a large number of patients after randomisation (Austin et al. 1993). It was stated that the four individual trials (Taverner 1954, May et al. 1976, Ünüvar et al. 1999, Lagalla et al. 2002) which met the inclusion criteria in the Cochrane review were too small (in all only including 179 patients) to detect moderate but important benefits with the use of corticosteroid therapy for Bell’s palsy (Salinas et al. 2004). None of the trials commented on side effects of the medical therapy. It was further concluded that there is a need for a randomised controlled trial with a large number of patients to determine reliably whether the use of corticosteroid therapy in patients with Bell’s palsy offers a real benefit (Salinas et al. 2004). Antiviral treatment: There is increasing evidence that the nerve injury in Bell’s palsy is caused by reactivation of viruses of the herpes group (McCormick 1972, Adour et al. 1975, Murakami et al. 1996, Furuta et al. 2000). The proposed link between Bell’s palsy and herpes viruses has led to the use of an antiviral agent (aciclovir or valaciclovir) in the treatment of the disease. Valaciclovir is a prodrug that is rapidly and nearly completely converted to aciclovir, and its bioavailability is three-fold to five-fold that of aciclovir (Beutner 1995). In 1996, Adour et al. (Adour et al. 1996) compared the outcome of 99 Bell’s palsy patients treated with aciclovir (400 mg 5 times daily for 10 days) in combination with prednisone (30 mg twice daily for five days gradually reduced to 5 mg twice daily during the next 5 days) versus placebo plus prednisone with a 4-month follow-up. Treatment with aciclovir plus prednisone was statistically more effective in returning volitional muscle 16.

(179) motion (recovery profile of 10; p=0.02) and in preventing nerve degeneration (p=0.05) than in patients treated with placebo plus prednisone (Adour et al. 1996). The prospective, controlled, randomised study of 101 patients by De Diego and co-workers (De Diego et al. 1998), reported that 47 patients treated with prednisone (1 mg/kg body weight for 10 days, then tapering over the next 6 days) had a better clinical recovery at 3 months than 54 patients treated with aciclovir (2,400 mg in three daily doses for 10 days). In 2000, Antunes et al. (Antunes et al. 2000) published a double-blind, randomised, placebocontrolled trial that included 46 patients receiving placebo or deflazocort (60 mg daily for 2 days, then tapering over the next 5 days) or deflazocort plus valaciclovir (1,500 mg daily for one week). No statistical difference in time to recovery of facial function among the three groups was found. In 2007, Hato et al. (Hato et al. 2007) presented a prospective, randomised, placebocontrolled study that compared a combination of valaciclovir (1,000 mg daily for 5 days) plus prednisolone (60 mg daily for 5 days, then tapering for a total treatment of 10 days) with placebo plus prednisolone. They reported a significantly higher recovery rate in the valaciclovir plus prednisolone group compared with placebo plus prednisolone. However, criticism has been voiced regarding this trial because it was not double-blinded and only 75% of the patients enrolled were analysed (Gilden and Tyler 2007). The review of Grogan et al. (Grogan and Gronseth 2001) included three studies using aciclovir to treat of Bell’s palsy (Ramos et al. 1992, Adour et al. 1996, De Diego et al. 1998). They concluded that aciclovir (combined with prednisone) is possibly effective in improving facial function in Bell’s palsy, but that further well-designed studies are needed (Grogan and Gronseth 2001). The Cochrane review by Allen and co-workers (Allen and Dunn 2004) included three controlled trials (Adour et al. 1996, De Diego et al. 1998, Antunes et al. 2000) involving aciclovir or valaciclovir in the treatment of Bell’s palsy. The authors concluded that more data are needed from a large multicentre, randomised, controlled, and blinded study with at least 12 months follow-up before a definitive recommendation can be made regarding the effect of aciclovir or valaciclovir (Allen and Dunn 2004). To clarify the effect of corticosteroids and anti-viral treatment in Bell’s palsy, two large randomised, double-blind, placebo-controlled trials were performed in Europe in the 2000’s; the Scandinavian Bell’s palsy trial testing prednisolone and valaciclovir (Engström et al. 2008) and a Scottish trial testing prednisolone and aciclovir (Sullivan et al. 2007). The recently published Scottish trial included 551 patients recruited within 72 hours of onset of palsy and randomly assigned to 10 days of treatment with prednisolone (25 mg twice daily), aciclovir (400 mg five times daily), both agents, or placebo. It was concluded that early treatment with prednisolone significantly improves the chances of complete recovery (House-Brackmann grade I) at 3 and 9 months, and that there was no benefit of aciclovir, neither as monotherapy nor in combination with prednisolone (Sullivan et al. 2007). 17.

(180) Pain in Bell’s palsy Ipsilateral pain around the ear, in the face and/or neck occurs in 33 to 70% of patients in the early stage of Bell’s palsy (Adour et al. 1972, May et al. 1976, Adour et al. 1978, Hydén et al. 1982, Katusic et al. 1986, Gavilan et al. 1988, Peitersen 2002). The aetiopathological background for this pain is unclear. It has, however, been reported that the presence of pain indicates a worse prognosis for facial recovery (Hydén et al. 1982, Katusic et al. 1986, Gavilan et al. 1988, Peitersen 2002). In contrast, other workers claim that pain has no prognostic value in the disease (May et al. 1976, Adour et al. 1978, Chida et al. 2002). The effect of prednisolone and/or valaciclovir on ipsilateral pain is controversial and needs further evaluation. Adour et al. stated that prednisone relieves pain in Bell’s palsy and that 15% of patients experience return of postauricular pain when prednisone is tapered (Adour et al. 1972). These results were not confirmed by Austin and co-workers who found no significant difference in pain between patients treated with prednisone or placebo (Austin et al. 1993).. Recovery rates in Bell’s palsy Earlier Bell’s palsy trials have reported diverging rates of facial recovery, both in treated and untreated patients. In the large non-controlled study on the spontaneous course of Bell’s palsy, including 1,701 cases studied over a period of 25 years, Peitersen reported that about 70% of the patients recovered without medical treatment within six months (Peitersen 2002). He concluded that no kind of treatment, including prednisone, was able to improve prognosis and stated that the time had now come to stop the use (or misuse) of prednisone (Peitersen 2002). In contrast, the world’s two largest (and recently published) randomised controlled Bell’s palsy trials both concluded that prednisolone does improve restitution of facial function, whereas antiviral treatment had no proven effect (Sullivan et al. 2007, Engström et al. 2008). Sullivan and coworkers reported complete recovery of facial function, defined as HouseBrackmann Grade I, in 90% of patients treated with prednisolone and in 75% in patients not receiving prednisolone (Sullivan et al. 2007). In the study of Engström et al., the corresponding proportions of patients with complete recovery, defined as Sunnybrook scale score of 100, were 72% and 57% (Engström et al. 2008). In two recent Japanese trials, in which the Yanagihara grading system was used to evaluate facial function, treatment with prednisolone resulted in recovery rates of 80% (Kawaguchi et al. 2007) and 90% (Hato et al. 2007). The studies mentioned above have important methodological differences that may affect the recovery rates, i.e. inclusion/exclusion criteria, facial grading systems, follow-up time, statistical methods and definitions of facial recovery. This impact of study design on recovery rates has recently been to the subject of discussion since studies of similar treatments report varying rates of recovery (Gilden and Tyler 2007, Davenport et al. 2008, Hato et al. 2008). 18.

(181) Aims of the study. • • • • • •. To assess the agreement between the Sunnybrook facial nerve grading system and the House-Brackmann and Yanagihara systems To study the short and long-term treatment effect of prednisolone and valaciclovir on facial nerve recovery in a large number of patients with Bell’s palsy To study the effect of prednisolone and valaciclovir on synkinesis, as well as side effects of the study drugs To evaluate the effect of prednisolone and valaciclovir on ipsilateral pain around the ear and in the face or neck in Bell’s palsy To study the incidence and intensity of pain during the first two months of palsy and to assess its prognostic value To evaluate the influence of different analysis methods and definition of facial recovery on recovery rates in Bell’s palsy. 19.

(182) Material and Methods. Patients Paper I From March 2001 to May 2003, 100 examinations were performed on 62 patients (30 women, 32 men) with varying degrees of peripheral facial palsy. The palsy was left/right-sided in equal numbers of the 62 patients. Their median age was 48 years (range 15–87) and median time between onset of palsy (day 1) and assessment was 50 days (average 129 days, range 1–4,000 days). Sixty patients were diagnosed as having Bell’s palsy (94 evaluations) and 2 herpes zoster oticus (six evaluations). Papers II–IV From May 2001 to September 2007, a prospective, randomised, doubleblind, placebo-controlled, multicentre trial with a 12-month follow-up was performed in patients with Bell’s palsy. Patients aged 18 to 75 years with onset of palsy within 72 hours were recruited in 16 public otorhinolaryngological centres in Sweden and one in Finland. Altogether, 1,953 patients (910 women, 1,043 men) with acute peripheral facial palsy were screened. At the initial examination, 1,114 of the 1,953 patients did not meet the inclusion criteria and were registered in separate forms. Accordingly, 839 of the 1,953 patients met the inclusion criteria and were randomised into four treatment arms. Ten randomised patients did not take any study drug and were excluded. Consequently, 829 patients (341 women, 488 men) were included in the modified intention-to-treat (ITT) analysis. Of these, 206 received placebo, 210 prednisolone, 207 valaciclovir, and 206 prednisolone plus valaciclovir. Thus, 416 of the patients received prednisolone and 413 did not. Similarly, 413 received valaciclovir while 416 were not treated with valaciclovir. Baseline characteristics of the four treatment groups were similar with regard to median age, gender, side of palsy, time from onset of palsy to treatment start, and the median Sunnybrook and House-Brackmann scores. During follow-up, the investigating physicians diagnosed diseases that may have been the cause of palsy; 67 borreliosis, 5 herpes zoster oticus, and 18 patients with other diseases (e.g. sarcoidosis, multiple sclerosis, cerebrovascular disease, and parotid tumour). These 90 patients were included in the ITT analysis. The 12-month follow-up visit was made by 743 of the 829 patients (90%). 20.

(183) Methods Paper I Case histories were obtained and the patients were examined by an otolaryngologist. They were asked to perform the facial movements at least three times and clinical grading was carried out according to the Sunnybrook (Ross et al. 1996), House-Brackmann (House and Brackmann 1985) and Yanagihara facial nerve grading systems (Yanagihara 1977). Thirty-six patients were examined once during the course of palsy, 19 patients twice, 3 three times, 3 four times and one patient five times. Hence, the total number of evaluations was 100. Three clinicians with special interest in facial nerve disorders conducted the study; two were consultants and one was a registrar. Papers II–IV Patients were assigned to one of four equally large treatment groups in a factorial fashion: placebo plus placebo; prednisolone 60 mg daily for five days, then tapering by 10 mg per day for the next five days plus placebo; valaciclovir 1,000 mg three times daily for seven days plus placebo; or prednisolone plus valaciclovir. Treatment started within 72 hours of onset of palsy. Facial function was assessed by Sunnybrook and House-Brackmann grading systems. Baseline assessment before treatment start included otorhinolaryngological examination, grading of facial function, measurement of ipsilateral pain, registration of concurrent medication, and serum analysis for Borrelia burgdorferi antibodies. Follow-up visits were scheduled for day 11 to 17, 1 month, 2 months, 3 months, 6 months, and 12 months after inclusion. If recovery was complete (Sunnybrook=100) at 2 or 3 months, the next follow-up was at 12 months. Otorhinolaryngological examination and registration of ipsilateral pain was performed at each visit during the first 2 months. The incidence and intensity of ipsilateral pain around the ear and in the face or neck was registered at baseline, day 11 to 17, 1 month, and 2 months. Patients graded the intensity of pain by marking a cross on a visual analogue scale (VAS) and they were not allowed to see their previous gradings. The VAS grading ranges from 0 to 10, where 0 represents no pain and 10 very severe pain. Convalescent serum for Borrelia burgdorferi antibodies was taken at 2 months.. 21.

(184) Statistical analysis Paper I To assess the agreement between the grading systems, we used weighted Kappa values (with 95% confidence intervals) of quintiles of the original scales. The interpretations of the numerical values of Kappa coefficients by Landis & Koch have become standard (Landis and Koch 1977). A Kappa value in the interval 0.00–0.20 means slight agreement, 0.21–0.40 fair, 0.41– 0.60 moderate, 0.61–0.80 substantial and 0.81–1.00 almost perfect agreement. The assessments were stratified into the time periods 1–14 days, 15– 60 days, 61–180 days and >180 days. For each period/stratum, Kappa values were calculated (95% confidence intervals) and the null hypothesis of equal Kappa values between strata was tested. Papers II and III The ITT principle was used for the analyses. All randomised patients receiving at least one dose of study medication were included in the analysis, but subjects who did not initiate therapy were excluded. Therefore, our ITT analysis should be considered a modified ITT analysis. Last-observationcarried-forward method was applied; missing data points were imputed in the post-baseline follow-up visits from the last observation available for each patient. An interaction test was performed on the primary endpoint to reveal any synergistic effect of the combination of prednisolone and valaciclovir. Results are given with continuous variables as median values with interquartile range (IQR; 25th to 75th percentiles) and dichotomous data as proportions with 95% confidence interval (CI), using the normal approximation approach. Kaplan-Meier methods were used to generate the survival curves. Categorical variables were compared by Fisher’s exact test. Continuous variable comparisons were performed using Kruskal-Wallis test. Cox proportional hazards models were used to estimate the hazard ratio (HR) of recovery, including 95% CI. The assumption for proportional hazards was tested using Schoenfeld residuals (p=0.73) (paper II). Paper IV Recovery rates of facial function were assessed with four different analysis methods: a) Recovery defined as Sunnybrook scale score of 100 at 12 months based on the intention-to-treat principle and last-observation-carried-forward method (n=829). This analysis is referred to as “last observation carried forward, Sunnybrook=100”. b) Recovery defined as House-Brackmann Grade I at 12 months based on the intention-to-treat principle and last-observation-carried-forward method (n=829). This analysis is referred to as “last observation carried forward, House-Brackmann=I”. 22.

(185) c) Recovery defined as House-Brackmann Grade I in patients followed until recovery plus those remaining at the 12-month follow-up (n=782) (47 patients were lost to follow-up before recovery and were excluded). This analysis is referred to as “complete-case analysis, House-Brackmann=I”. d) Recovery defined as House-Brackmann Grade II in patients followed until recovery plus those remaining at the 12-month follow-up (n=797) (32 patients were lost to follow-up before recovery and were excluded in this analysis). This analysis is referred to as “complete-case analysis, House-Brackmann II”.. 23.

(186) Results. Paper I Agreement between the Sunnybrook, House-Brackmann and Yanagihara facial nerve grading systems in Bell’s palsy Overall agreement between the Sunnybrook, House-Brackmann and Yanagihara facial grading systems When comparing the overall agreement between the Sunnybrook and the House-Brackmann facial grading system in 100 evaluations, the weighted Kappa value was 0.59 (Spearman correlation coefficient=0.76) (Table 5). Patients graded as House-Brackmann II had Sunnybrook composite scores ranging from 53 to 87 (scoring interval=35), House-Brackmann III was Sunnybrook gradings 25 to 64 (scoring interval=40), House-Brackmann IV corresponded to Sunnybrook gradings 9 to 55 (grading interval=47), HouseBrackmann V was Sunnybrook gradings 4 to 42 (interval=39), and patients with House-Brackmann VI were graded 8 to 17 (interval=10) in the Sunnybrook system. The best overall agreement was demonstrated between the Sunnybrook and Yanagihara grading systems with a weighted Kappa value of 0.72 (Spearman correlation coefficient=0.84) (Table 5). If a direct conversion (Sunnybrook=Yanagihara x 2.5) was made, these results would indicate that Sunnybrook gradings generally score lower than the corresponding values in the Yanagihara system, both in the interval with low scoring values (severe palsy) and the interval with high values (mild palsy). Weighted Kappa was 0.64 (Spearman correlation coefficient=0.75) when comparing the overall agreement between House-Brackmann and Yanagihara scores (Table 5). If the conversion table suggested by Satoh et al. (Satoh et al. 2000) is used, the number of individual Yanagihara gradings within the converted House-Brackmann percentage intervals will be 41 (41%). Agreement between the Sunnybrook, House-Brackmann and Yanagihara facial grading systems during the course of palsy days 1–14, 15–60, 61–180 and after day 180 The agreement between the four grading systems was analysed for the intervals: days 1–14 (38 evaluations), days 15–60 (28 evaluations), days 61–180 (20 evaluations) and after day 180 (14 evaluations) (Table 5). The congruence values of Sunnybrook vs. House-Brackmann scores, measured by 24.

(187) weighted Kappa for the four time intervals, were: 0.56, days 1–14; 0.53, days 15–60; 0.71, days 61–180; and 0.55, after day 180. The corresponding Kappa values for Sunnybrook vs. Yanagihara scores, were 0.76, 0.74, 0.71 and 0.38, respectively. The weighted Kappa scores for House-Brackmann vs. Yanagihara grades were: 0.64, days 1–14; 0.66, days 15–60; 0.74 days 61– 180; and 0.37 after day 180. Table 5. Overall agreement in 100 evaluations and agreement over time between the Sunnybrook, House-Brackmann and Yanagihara facial grading systems in Bell’s palsy (n = 94) and herpes zoster oticus (n = 6).. Sunnybrook vs. House–Brackmann Sunnybrook vs. Yanagihara House-Brackmann vs. Yanagihara Total between all grading systems. Weighted Spearman Kappa correlation coefficient values n = 100 n = 100 0.59 0.76. Days 1-14 Weighted Kappa n = 38 0.56. Days 15-60 Weighted Kappa n = 28 0.53. Days 61-180 Weighted Kappa n = 20 0.71. Days181Weighted Kappa n = 14 0.55. 0.72. 0.84. 0.76. 0.74. 0.71. 0.38. 0.64. 0.75. 0.64. 0.66. 0.74. 0.37. 0.65. 0.78. 0.65. 0.64. 0.72. 0.43. Paper II Prednisolone and valaciclovir in Bell's palsy: a randomised, doubleblind, placebo-controlled, multicentre trial A multivariate Cox analysis showed that time to complete recovery (Sunnybrook=100) was significantly shorter in the 416 patients receiving prednisolone compared with the 413 not treated with prednisolone (HR 1.40; 95% CI 1.18 to 1.64; p<0.0001). Time to recovery did not differ between the 413 patients who received valaciclovir and the 416 who did not receive valaciclovir (HR 1.01; 95% CI 0.85 to 1.19; p=0.90). Interaction between the effects of prednisolone and valaciclovir was not found (p=0.59). The median time to complete recovery was 75 days in the prednisolone plus placebo arm, which was less than the 104 days in placebo plus placebo (p=0.04) and 135 days in valaciclovir plus placebo (p=0.03) (Figure 1). Patients receiving prednisolone had significantly higher recovery rates at 3, 6, and 12 months than no-prednisolone patients (Table 6 and 7). At 12 months, 300 (72%) of patients in the prednisolone group (n=416) had recovered compared with 237 (57 %) in no-prednisolone (n=413) (p<0.0001). The outcome for the 413 patients treated with valaciclovir did not differ from the 416 in the no-valaciclovir group (p=0.66) (Table 7). Of the 743 patients with a 12-month follow-up, synkinesis was present in 51 of 370 (13.8%) patients treated with prednisolone compared with 107 of 25.

(188) 373 (28.7%) in those who did not receive prednisolone (difference -14.9%, 95% CI -20.7 to -9.1; p<0.0001). In the 369 patients receiving valaciclovir, 73 (19.8%) had synkinesis compared with 85 of 374 (22.7%) in the novalaciclovir group (difference -2.9%, 95% CI -8.8 to 2.9; p=0.37). Non-serious adverse events were reported by 92 patients; 25 in the placebo plus placebo arm, 21 in prednisolone plus placebo, 19 in valaciclovir plus placebo, and 27 in patients who received prednisolone plus valaciclovir. There was no difference in the number of patients with adverse events (Fisher’s exact test) between prednisolone plus placebo and placebo plus placebo (difference -2.1%, 95% CI -8.3 to 4.1; p=0.53), valaciclovir plus placebo and placebo plus placebo (difference -3.2%, 95% CI -9.2 to 2.9; p=0.34), or prednisolone plus valaciclovir and placebo plus placebo (difference 0.9%, 95% CI -5.7 to 7.5; p=0.88). 100. Num bers at risk (tim e points) t1 194 200 202 200. Complete Recovery (%). 80. t2 180 184 181 174. t3 143 130 138 123. t4 113 99 115 94. t5 98 73 102 77. t6 77 57 84 54. (PL) (P) (V) (PV). 60 50 40. 20 Placebo (PL) Prednisolone (P) Valacyclovir (V) Prednisolone plus valacyclovir (PV). 0 0. 100 Md Md PV P. Md PL. 200 Md V. 300. 400. Days. Figure 1. Kaplan-Meier estimates of patients with complete recovery (Sunnybrook=100) in the four treatment groups (n=829). The median time (Md) to complete recovery for each treatment group is also illustrated. Numbers at risk; t1=day 11-17, t2=1 month, t3=2 months, t4=3 months, t5=6 months, and t6=12 months.. 26.

(189) 65, 31.0 (24.7–37.3). 60, 29.0 (22.8–35.2). 71, 34.5 (27.9–41.0). 12, 5.8 (2.6–9.0). 15, 7.3 (3.7–10.9). Sunnybrook=100. House-Brackmann=I. 78, 37.9 (31.2–44.5). 65, 31.4 (25.0–37.8). 13, 6.3 (3.0–9.6) Prednisolone plus valaciclovir (n=206). 12, 5.8 (2.6–9.0). Sunnybrook=100. House-Brackmann=I. 71, 33.8 (27.4–40.3). 14, 6.7 (3.3–10.1). 18, 8.6 (4.8–12.4) Valaciclovir plus placebo (n=207). House-Brackmann=I. Sunnybrook=100. 56, 27.2 (21.1–33.3). 63, 30.6 (24.2–36.9). 16, 7.8 (4.1–11.4). At 1 month n, % (95% CI). 16, 7.8 (4.1–11.4) Prednisolone plus placebo (n=210). House-Brackmann=I. Sunnybrook=100. n, % (95% CI) Placebo plus placebo (n=206). At 11-17 days. 116, 56.3 (49.5–63.1). 109, 52.9 (46.0–59.8). 94, 45.4 (38.6–52.2). 89, 43.0 (36.1–49.8). 111, 52.9 (46.1–59.7). 107, 51.0 (44.1–57.8). 94, 45.6 (38.8–52.5). 88, 42.7 (35.9–49.5). n, % (95% CI). At 2 months. 134, 65.0 (58.5–71.6). 124, 60.2 (53.4–66.9). 113, 54.6 (47.8–61.4). 104, 50.2 (43.4–57.1). 137, 65.2 (58.7–71.7). 135, 64.3 (57.8–70.8). 111, 53.9 (47.0–60.8). 105, 51.0 (44.1–57.9). n, % (95% CI). At 3 months. 149, 72.3 (66.2–78.5). 141, 68.5 (62.1–74.9). 120, 58.0 (51.2–64.8). 111, 53.6 (46.8–60.5). 150, 71.4 (65.3–77.6). 143, 68.1 (61.7–74.5). 127, 61.7 (54.9–68.4). 117, 56.8 (50.0–63.6). n, % (95% CI). At 6 months. 164, 79.6 (74.1–85.2). 152, 73.8 (67.7–79.8). 133, 64.3 (57.7–70.8). 119, 57.5 (50.7–64.3). 160, 76.2 (70.4–82.0). 148, 70.5 (64.3–76.7). 133, 64.6 (58.0–71.1). 118, 57.3 (50.5–64.1). n, % (95% CI). At 12 months. Table 6. The proportion of patients in the intention-to-treat analysis with complete recovery defined as Sunnybrook=100 or HouseBrackmann=I at days 11-17, 1 month, and 2, 3, 6, and 12 months. Last-observation-carried-forward was applied for missing data points. Data are number of patients (n), % in each treatment group with complete recovery (95% CI)..

(190) 14%. vs 104, 50% (43–57) (5–24). 135, 64% (58–71). Valaciclovir (n=413) vs no Valaciclovir (n=416). Prednisolone plus Placebo (n=210) vs Valaciclovir plus Placebo (n=207). 228, 55% (50–60) vs 240, 58% (53–63). Prednisolone (n=416) vs no Prednisolone (n=413) -3% (-9–4). At 3 months n Diff 259, 62% (58–67) 12% vs 209, 51% (46–55) (5–18). 0.004. 0.48. 15%. -2% (-8–5). vs 111, 54% (47–61) (5–24). 143, 68% (62–75). 252, 61% (56–66) vs 260, 63% (58–67). At 6 months P n Diff 0.0007 284, 68% (64–73) 13% vs 228, 55% (50–60) (7–20). 0.003. 0.068. 13%. 2% (-5–8). vs 119, 58% (51–64) (4–22). 148, 71% (64–77). 271, 66% (61–70) vs 266, 64% (59–69). At 12 months P n Diff 0.0001 300, 72% (68–76) 15% vs 237, 57% (53–62) (8–21). 0.006. 0.66. P <0.0001. Table 7. Rates of complete recovery (Sunnybrook=100). Data are number (n) of patients, % for each treatment modality with complete recovery (95% CI) and differences in % (95% CI)..

(191) Paper III The course of pain in Bell’s palsy: treatment with prednisolone and valaciclovir At baseline (within 72 hours of onset of palsy), 412 of 826 patients (50%) in the ITT-analysis reported ipsilateral pain around the ear, in the face or in the neck. The incidence of pain in this acute stage was similar in all four treatment arms (Table 8). The median intensity of pain measured with VAS in the treatment arms is also shown in Table 8. At the first follow-up (day 11 to 17), 369 of 807 patients (46%) experienced pain (Table 8). Of these 369 patients, 118 (32%) reported pain for the first time, i.e. they had not experienced pain at baseline. Intensity of pain was similar between the arms (Table 8). Factorial analysis showed no significant difference in incidence or intensity of pain between the 403 patients treated with prednisolone compared with the 404 patients not treated with prednisolone (p=0.57, p=0.58 respectively), or between the 407 patients receiving valaciclovir compared with the 400 patients not receiving this medication (p=0.72, p=0.08 respectively). At 1 month, 128 of 717 patients (18%) exhibited ipsilateral pain around the ear, in the face or the neck (Table 8). Six of these 128 (5%) reported pain for the first time. The intensity of pain was similar between the four arms at 1 month (Table 8). Treatment with prednisolone or valaciclovir did not affect the incidence (p=0.92, p=0.38 respectively) or intensity of pain (p=0.37, p=0.37 respectively) at 1 month (factorial analysis). Two months after palsy onset, 53 of 637 patients (8%) reported pain (Table 8). Of these 53 patients, six (11%) experienced pain for the first time. As in the earlier follow-ups, the incidence and intensity of pain was similar among the four treatment arms (Table 8). Furthermore, treatment with prednisolone or valaciclovir did not influence incidence (p=0.89, p=0.77 respectively) or intensity of pain (p=0.85, p=0.69 respectively). During the first 2 months of palsy, 542 of 829 patients (65%) reported pain at some point; 133 of 206 (65%) receiving placebo plus placebo, 140 of 210 (67%) receiving prednisolone plus placebo, 145 of 207 (70%) receiving valaciclovir plus placebo, and 124 of 206 (60%) receiving prednisolone plus valaciclovir. Table 9 presents recovery rates for facial motor function at 12 months in relation to pain in the first two weeks of palsy as determined by factorial analysis. There was no significant difference in recovery rates at 12 months within any of the treatment groups in relation to pain within 72 hours (Table 9). In patients reporting pain at day 11 to 17, 204 of 369 (55%) had complete recovery at 12 months in comparison with 323 of 438 (74%) without pain at this follow-up (p<0.0001) (Table 9). Significant differences in recovery rates. 29.

(192) 38 (25–56) 4 (3–5) 107/210 (51%) 3 (2–5) 96/201 (48%) 3 (2–5) 39/174 (22%) 2 (1–4) 13/159 (8%) 2 (1–3). 84/199 (42%) 3 (1–5). 29/179 (16%) 2 (1–3) 15/162 (9%) 1 (1–2). 82 (39%) 128 (61%) 40 (30–52). 38 (23–51) 4 (3–5) 93/204 (46%) 2 (1–5). 93 (45%) 113 (55%) 39 (30–53). Prednisolone plus Placebo n=210. 13/165 (8%) 1 (1–3). 37/186 (20%) 1 (1–3). 105/205 (51%) 3 (2–5). 41 (21–54) 4 (3–5) 111/206 (54%) 3 (1–5). 86 (42%) 121 (58%) 40 (32–54). Valaciclovir plus Placebo n=207. 12/151 (8%) 1 (0–2). 23/178 (13%) 2 (1–3). 84/202 (42%) 3 (2–5). 38 (23–54) 4 (3–5) 101/206 (49%) 2 (1–5). 80 (39%) 126 (61%) 42 (31–56). Prednisolone plus Valaciclovir n=206. — Indicates data not presented; IQR, interquartile range; VAS, visual analogue scale. a Data missing for three patients,b Data missing for 22 patients,c Data missing for 122 patients,d Data missing for 192 patients. Within 72 hours Facial nerve grading score Sunnybrook, median (IQR) House-Brackmann, median (IQR) Ipsilateral pain, n (%)a VAS, median (IQR) 11–17 days Ipsilateral pain, n (%)b VAS, median (IQR) 1 Month Ipsilateral pain, n (%)c VAS, median (IQR) 2 Months Ipsilateral pain, n (%)d VAS, median (IQR). Sex Female, n (%) Male, n (%) Age, median (IQR). Placebo plus Placebo n=206. Table 8. Characteristics of the patients in the four treatment arms. 39 (23–54) 4 (3–5) 412/826 (50%) 2 (1–5). 341 (41%) 488 (59%) 40 (31–54). n=829. Total. 0.98 0.85. 53/637 (8%) 1 (1–3). 0.10 128/717 (18%) 0.70 2 (1–3). 0.16 369/807 (46%) 0.36 3 (2–5). — — — —. — — —. p.

(193) 141/212, 66.5% (60.1–72.9) 119/200, 59.5% (52.6–66.4). Valaciclovir (n=413) No Valaciclovir (n=416). 130/200, 65.0% (58.3–71.7) 146/214, 68.2% (61.9–74.5). 151/208, 72.6% (66.5–78.7) 125/206, 60.7% (54.0–67.4). 276/414, 66.7% (62.1–71.2). Data missing for three patients,b Data missing for 22 patients. 149/208, 71.6% (65.5–77.8) 111/204, 54.4% (47.5–61.3). Prednisolone (n=416) No Prednisolone (n=413). a. 260/412, 63.1% (58.4–67.8). All patients (n=829). Complete recovery at 12 monthsa (n, % [95% CI]) Pain within 72 hours No pain within 72 hours (n=412) (n=414). 0.76 0.07. 0.91 0.23. 0.31. p. 103/189, 54.5% (47.3–61.7) 101/180, 56.1% (48.8–63.4). 117/180, 65.0% (58.0–72.0) 87/189, 46.0% (38.9–53.2). 204/369, 55.3% (50.2–60.4). 167/218, 76.6% (70.9–82.3) 156/220, 70.9% (64.9–77.0). 177/223, 79.4% (74.0–84.7) 146/215, 67.9 % (61.6–74.2). 323/438, 73.7% (69.6–77.9). Complete recovery at 12 monthsb (n, % [95% CI]) Pain at day 11 to 17 No pain at day 11 to 17 (n=369) (n=438). Table 9. Complete recovery rates (Sunnybrook=100) at 12 months in relation to incidence of pain within 72 hours and at day 11 to 17. <0.0001 0.0024. 0.0015 <0.0001. <0.0001. p.

(194) at 12 months in relation to pain vs. no pain at day 11 to 17 were found in the prednisolone (65% vs. 79%, p=0.0015), no-prednisolone (46% vs. 68%, p=<0.0001), valaciclovir (55% vs. 77%, p=<0.0001), and no-valaciclovir groups (56% vs. 71%, p=0.0024) (Table 9). Time to facial recovery was shorter in patients with no pain compared with pain at day 11 to 17 (p<0.0001). During the first month, painkillers were taken by 36 of 206 patients (17%) in the placebo plus placebo arm, 28 of 210 (13%) in prednisolone plus placebo, 41 of 207 (20%) in valaciclovir plus placebo, and 32 of 206 (16%) in prednisolone plus valaciclovir (p=0.33). The most common painkillers were non-steroidal anti-inflammatory drugs followed by paracetamol.. Paper IV The effect of study design and analysis methods on recovery rates in Bell’s palsy a) Recovery rates for last observation carried forward, Sunnybrook=100 (n=829) With recovery defined as Sunnybrook scale score of 100, analysis of the 829 patients based on the intention-to-treat principle and last-observationcarried-forward method showed that 300 of the 416 patients (72%) in the prednisolone group had recovered at 12 months compared with 237 of 413 patients (57%) who did not receive prednisolone (Table 10 and Figure 2). For valaciclovir, the corresponding values at 12 months were 271 of 413 (66%) for the valaciclovir group and 266 of 416 (64%) in the no-valaciclovir group (Table 10). b) Recovery rates for last observation carried forward, House-Brackmann=I (n=829) When House-Brackmann Grade I was defined as recovery, the recovery rates at 12 months in the 829 intention-to-treat patients were 324 of 416 (78%) for those receiving prednisolone and 266 of 413 (64%) for patients not receiving prednisolone (Table 10 and Figure 2). In the valaciclovir group, 297 of 413 (72%) had recovered, while in patients not treated with valaciclovir this figure was 293 of 416 (70%) (Table 10). c) Recovery rates in complete-case analysis, House-Brackmann=I (n=782) With recovery defined as House-Brackmann Grade I in patients followed until recovery (39 patients had House-Brackmann Grade I before 12-months) plus the remaining patients with a 12-month follow-up (n=782), 335 of 389 patients (86%) in the prednisolone group had recovered compared with 277 of 393 (70%) in the no-prednisolone group (Table 10 and Figure 2). The corresponding rates for valaciclovir/no-valaciclovir groups were 307 of 386 (80%) and 305 of 396 (77%) respectively (Table 10). 32.

(195) d) Recovery rates in complete-case analysis, House-BrackmannII (n=797) When definition of recovery was House-Brackmann Grade II in patients followed until recovery plus the remaining patients with a 12-month follow-up (n=797), recovery rates were 380 of 396 (96%) in prednisolone-treated patients and 353 of 401 (88%) in those not treated with prednisolone (Table 10 and Figure 2). The corresponding recovery in the valaciclovir group was 363 of 393 (92%) and in the no-valaciclovir group, 370 of 404 (92%) (Table 10).. Figure 2. Recovery rates in Bell’s palsy patients receiving prednisolone (n=416) and not receiving prednisolone (n=413). Last-observation-carried-forward method with recovery defined as Sunnybrook scale score of 100 and House-Brackmann Grade I, and complete-case analysis method with recovery defined as House-Brackmann Grade I and Grade II respectively.. 33.

(196) a) Last observation carried forward, Sunnybrook=100, n=829 b) Last observation carried forward, House-Brackmann=I, n=829 c) Complete-case analysis, House-Brackmann=I, n=782 d) Complete-case analysis, House-Brackmann≤II, n=797 237/413, 57% (53–62) 266/413, 64% (60–69) 277/393, 70% (66–75) 353/401, 88% (85–91). 324/416, 78% (74–82). 335/389, 86% (83–90). 380/396, 96% (94–98). No prednisolone n, % (95% CI). 300/416, 72% (68–76). Prednisolone n, % (95% CI) Valaciclovir n, % (95% CI). No valaciclovir n, % (95% CI). <0.0001 363/393, 92% (90–95) 370/404, 92% (89–94). <0.0001 307/386, 80% (75–84) 305/396, 77% (73–81). <0.0001 297/413, 72% (68–76) 293/416, 70% (66–75). <0.0001 271/413, 66% (61–70) 266/416, 64% (59–69). p value. Table 10. Recovery rates in the treatment groups according to the different methods of analysis and definitions of recovery.. 0.70. 0.44. 0.65. 0.66. p value.

(197) Discussion. In the large Scandinavian Bell’s palsy trial including 829 patients (Paper II), we found that patients who received prednisolone had a shorter time to complete recovery of facial function and a more favourable outcome at 12 months than patients who did not receive prednisolone. Furthermore, synkinesis was less common in patients treated with prednisolone. Valaciclovir was not proven effective and not found to add effect to prednisolone. Similar results of corticosteroid and antiviral treatment were reported in a recent Scottish trial including 551 patients randomly assigned to 10 days of prednisolone 25 mg twice daily, aciclovir 400 mg five times daily, both agents, or placebo (Sullivan et al. 2007). The investigators of that trial concluded that early treatment with prednisolone significantly improved the chances of complete recovery at 3 and 9 months, while there was no benefit of aciclovir. The effect of prednisolone indicates that inflammation and edema of the facial nerve is part of the pathogenesis in Bell’s palsy. This agrees with previous findings in per-operative (Fisch and Esslen 1972) and MRI studies (Schwaber et al. 1990, Engström et al. 1993). The anti-inflammatory effect of prednisolone may reduce neural and perineural edema, which in turn affects motor function. Our study is the first large, placebo-controlled Bell’s palsy study testing valaciclovir alone against placebo. The bioavailability of valaciclovir is three-fold to five-fold that of aciclovir, implying a much higher anti-viral activity against herpes viruses (Lycke et al. 2003). We gave valaciclovir orally, 1,000 mg three times daily for seven days, providing a concentration well above the inhibitory level for herpes simplex virus but maybe only partially inhibitory for varicella zoster virus (Lycke et al. 2003). The ineffectiveness of valaciclovir may also be due to virus replication having declined before treatment start (Stjernquist-Desatnik et al. 2006), or that herpes viruses are not the main cause of Bell’s palsy. Prednisolone plus valaciclovir was not more effective on facial motor recovery than prednisolone alone (Figure 1). This is in accordance with previous findings using valaciclovir (Kawaguchi et al. 2007) or aciclovir (De Diego et al. 1998, Sullivan et al. 2007) plus corticosteroids. Nevertheless, an additional effect of aciclovir (Adour et al. 1996) or valaciclovir (Hato et al. 2007) on corticosteroids has been reported. However, the Adour trial (Adour et al. 1996) only followed patients for four months and the Hato study (Hato. 35.

No results found