Aims: Burning mouth syndrome (BMS) is characterized by a burning oral dysesthesia that cannot be attributed to other local or systemic disease entities. Clonazepam is a medication typically prescribed for the management of epilepsy, but has also demonstrated some success in the management of BMS. The purpose of this retrospective study was to evaluate the efficacy of a mode of clonazepam administration in which patients were asked to dissolve clonazepam slowly orally before swallowing for the management of pain associated with BMS.
Methods: Screening of clinical records obtained from January 2006 to June 2009 revealed 54 patients (10 men, 44 women, aged 56.98 ± 0.24 years) who met the criteria for inclusion in this study based on a clear diagnosis of BMS, and adherence to the clonazepam protocol for review. All reviewed patients were prescribed clonazepam doses between 0.5mg and 4.0mg daily. Pain was assessed quantitatively on an 11-point numerical scale (0 to 10) and/or qualitatively through written clinical records.
Results: The mean duration of treatment was 25 ± 3.29 weeks. The mean pain score reduction where patients gave a pre-treatment and final appointment pain score (n = 36) was 4.47 ± 0.43 points. 83% of patients (n = 45) obtained at least partial resolution of their BMS symptoms, and 67% (n = 36) obtained marked relief over the treatment period. Side effects were mostly transient and mild, however 7 patients withdrew from treatment due to adverse effects. The serum concentration of clonazepam in patients who obtained pain relief ranged from <10 µg/L to 30 µg/L, however no concentration-response relationship was observed.
KEY WORDS: burning mouth syndrome, oral, pain, clonazepam, lozenge.
Burning Mouth Syndrome (BMS) is a condition that is diagnosed based on the presence of an oral burning or similar dysesthesia in the absence of other dental or medical causes.1 Prevalence estimates range from 0.7% to 15%,2,3 a variation that is likely to be attributed to controversy surrounding the diagnostic criteria for this condition.4 The mean age of patients with BMS is around 60 years,5,6 and while some males are affected, the condition predominantly affects females.4,7 BMS may be accompanied by other oral symptoms including perceived xerostomia and taste disturbances,5,8 and this orofacial pain condition has also been linked to anxiety, depression and personality disorders in affected patients.1
A systematic review of the available literature in 2005 found that trials using antidepressants, analgesics and hormone replacement therapy produced no relief of symptoms associated with BMS.1 Of the trials that did demonstrate reduction of symptoms (alpha-lipoic acid, cognitive behavioral therapy, clonazepam), only one randomized controlled trial (RCT) by Gremeau-Richard et al.4 that assessed the topical administration of clonazepam for the treatment of BMS was deemed to be of sufficient design quality to provide sound conclusions.1
Clonazepam (Paxam- Alphapharm Pty Ltd.) is an anticonvulsant medication that is typically used for the management of epilepsy at adult doses of 4mg to 8mg per day,9 however this medication has also demonstrated another potential use at lower doses for the management of orofacial pain.4,10,11 In addition to the 2004 topical clonazepam RCT,4 two open-label trials have demonstrated reduction of BMS symptoms using clonazepam. In 1998, Grushka et al.12 found that a clonazepam tablet dose swallowed orally (systemically delivered) was effective in reducing symptoms associated with BMS. In the same year, Woda et al.13 reported that sucking a clonazepam tablet for 3 minutes then expectorating excess saliva was also effective.
While existing research into the use of clonazepam for BMS symptom reduction shows that independently, systemic administration and topical administration may each be valid modes of drug delivery,4,12,13 there is currently no research that examines the effects of a combined systemic and topical administration mode in which clonazepam tablets are dissolved orally before being swallowed. This study seeks to determine the efficacy of this ‘lozenge’ administration method through retrospective assessment of the treatment outcomes of BMS patients treated with this approach.
2.1 Patients and selection criteria
Patients from two centers (private Oral Medicine specialist practices in Brisbane and Sydney, Australia) consented at the time of initial consultation for their clinical records to be used for the purposes of teaching, learning and further research, with anonymity assured in accordance with normal protocol for all patients visiting these centers.
A key-word search of the clinical records databases from January 2006 to June 2009 was performed. Patients who met the inclusion criteria were diagnosed with BMS during this period, did not have concurrent local or systemic conditions to which their oral dysesthesia could be attributed (such as candidiasis or iron deficiency), attended at least 1 review appointment and began taking orally dissolved then swallowed clonazepam for the treatment of their condition. 54 patients met these inclusion criteria.
De-identified information collected from clinical records as part of this study included age, gender, referral information, description and location of dysesthesia(s), events associated with the onset of the condition, concurrent oral symptoms, medical history information, past treatments for the discomfort and all appointment details including appointment dates, dosage regimen, pain and side effect assessments. All available blood sample results were also reviewed. This protocol was evaluated and approved by the University of Queensland Medical Research Ethics Review Committee.
2.2 Dosage regimen
Patients within this study were prescribed doses of between 0.5mg and 4mg of clonazepam daily. Patients were instructed to take their clonazepam dose only after food and were advised not to eat or drink for at least 30 minutes afterwards. All patients were advised to dissolve clonazepam orally as a lozenge before swallowing. The clonazepam dose was initially escalated slowly. Patients began by dissolving only one 0.5mg tablet in the mouth nightly for the first week, then twice daily (morning and night) for the second week, then three times daily (morning, afternoon, and night) for the third week with subsequent review and alteration of dose where necessary. Most patients were prescribed a dose of 1.5mg of clonazepam daily in three divided doses, however dose was varied at the practitioner’s discretion according to individual patient requirements. Generally, where patients experienced side effects or were achieving relief and were eager to reduce their reliance on the medication, the dose was reduced. If patients were tolerating the medication well but not achieving complete resolution of symptoms, the dose was escalated.
2.3 Outcome assessment
Most patients were asked to verbally assign a pain score out of 10 (with 0 being no pain and 10 the worst pain imaginable) before beginning clonazepam therapy. Pain scores were not recorded at all appointments, however most patients were asked to assess pain throughout their treatment on the same scale to account for the average pain they had experienced in the period between visits. Where pain scores were not documented, qualitative assessments recorded in clinical charts were used. An attempt was made to gain an overall impression of patient response by combining qualitative and quantitative outcome assessments (Table 1). 50% improvement was set as a discriminator of partial versus marked pain reduction to meet common chronic pain resolution aims,1 and afford ease of comparison with existing quality research in this field.4
2.4 Blood sampling
14 patients had blood tests that included measurement of serum clonazepam levels during their treatment. Serum clonazepam readings were sought where blood sampling was required for other reasons such as the exclusion of systemic disease or blood monitoring for concurrent medical conditions. The period between the last dose of clonazepam and the time at which the blood sample was taken was recorded in most cases, however there was no standardization of this interval between patients or across separate test periods for the same patient.
2.5 Data Analysis
The efficacy of this mode of administration was primarily assessed based on the difference in pain from the first to the final appointment, although changes noted at the first review were also tested for significance using a two-tailed Mann-Whitney analysis (P < 0.05). Patients were categorized into four distinct groups according to their pain changes; (1) No relief obtained, (2) Partial reduction, (3) Marked reduction and (4) Complete resolution. A fifth category included patients who withdrew within the first week of treatment due to side effects and did not assess their pain changes.
GraphPad Prism 5.0 software was used to analyze the data obtained. A nonparametric (Spearman) two-tailed correlation matrix with 95% CI was used to assess the predictive power of age, gender, number of practitioners consulted before presentation, symptom duration before presentation, number of medications concurrently taken, number of psychoactive or central nervous system (CNS) depressant medications concurrently taken and pre-treatment pain severity on the overall pain resolution achieved. A two-tailed Mann-Whitney test was also used to detect significant differences in the presenting characteristics of separate response groups. Values of P less than 0.05 were considered statistically significant. All values are expressed as Mean ± SEM unless otherwise specified.
3.1 Population characteristics
Of the 54 patients who met the inclusion criteria of this study, 44 were female and 10 were male, with a mean age of 56.98 ± 0.24. Patients were most commonly referred for specialist management by their general dentist (n = 34) or medical practitioner (n = 9). Pain duration before presentation ranged from 1 month to 12.5 years, however most patients had been suffering for over 1 year (14.5 ± 3.44 months) before presenting to the specialist clinics. 24 patients experienced pain in multiple sites. The tongue was the most common site affected (n = 42) and 21 patients specified the anterior aspect of the tongue specifically as the source of their pain. Other sites frequently affected included the labial mucosa (n = 15) and the palate (n = 13). 25 patients reported perceived xerostomia (n = 16) and/or taste disturbances (n = 14) accompanying their oral pain at the time of presentation.
35 patients associated a specific event with the onset of their pain. The most common of these were family stress (n = 11), dental procedures (n = 11), general stress (n = 7), and illness or death of a family member (n = 6). 28 patients had attempted other treatment previously to address their oral pain. The most common of these were topical antifungal medications (n = 10), dental treatment (n = 7), mouthwashes (including Betadine, bicarbonate of soda, Biotene, Curasept, Dentyl, Listerine, Neutrafluor, peroxide and salt water) (n = 6), topical corticosteroids (n = 4) and topical anesthetics (n = 4).
19 patients noted anxiety, depression or significant stress as part of their medical history. 20 patients were taking psychoactive medications (sedatives, hypnotics, antidepressants or anti-anxiety agents) and 3 patients were taking other central nervous system (CNS) depressant medications (movement disorder medication, narcotic analgesics or anti-migraine preparations) during the treatment period. 8 patients were taking more than one psychoactive or CNS depressant medication concurrently.
3.2 Treatment duration and dosage regimen
The mean duration from the first appointment to the final appointment was 25 ± 3.29 weeks (range, 1 week to 119 weeks) with 3 ± 0.24 appointments per patient on average. Daily dosage of clonazepam ranged from 0.5mg to 4.0mg per day (Table 2). The median dose of clonazepam used was 1.5mg, which was most commonly divided into three 0.5mg doses each day.
3.3 Pain evaluation
The mean pre-treatment pain score recorded (n = 42) was 6.77 ± 0.25 points out of 10. Other key pain scores are summarized in Table 3. Where patients gave a score before treatment and at their first review appointment (7.20 +/- 1.01 weeks after the initial appointment), the mean pain score reduction was 4.09 +/- 0.40 points (n = 38, mean 4, range 0 to 10). The mean reduction observed where patients gave a pre-treatment and final appointment pain score (n = 36) was 4.47 ± 0.43 points (median 5, range 0 to 10). Where pain reduction could be evaluated as a percentage reduction from pre-treatment levels to final appointment levels (n = 37), the average reduction in pain was 69.38% ± 4.65% (median 71.43%, range 0 to 100%).
Using the categorization methods outlined (Table 1), 6 patients did not obtain relief from clonazepam, 9 patients obtained partial relief of symptoms, 21 patients obtained marked resolution of pain, 15 patients had complete resolution of their pain, and 3 patients withdrew within ten days of beginning clonazepam, before they made an assessment of pain resolution (Figure 1). 83.33% of patients (n = 45) obtained at least partial resolution of their BMS symptoms, and 66.67% (n = 36) obtained at least 50% or marked relief from taking clonazepam. This figure decreases to 62.96% when the two patients who had at least marked pain relief, but withdrew later during therapy due to side effects are removed from results regarding resolution.
A two-tailed correlation matrix did not show a significant correlation between gender, number of practitioners consulted before presentation, symptom duration before presentation, number of medications concurrently taken, number of psychoactive/CNS depressant medications concurrently taken and pre-treatment pain severity on the overall pain resolution achieved. There was a weak but insignificant association between lower patient age and improved treatment outcomes (P= 0.056). Comparison of patient groups according to treatment response (Table 4) did not reveal any significant differences in presenting characteristics, with the exception that patients with complete pain resolution were significantly younger than patients with only partial pain reduction (P = 0.007).
3.5 Side effects and withdrawal
37% of patients (n = 20) noted side effects during the course of treatment. For 13 patients these were only transient and/or mild and included drowsiness, tiredness, dizziness and changes in mood. 3 patients withdrew within 10 days of starting clonazepam due to side effects. 2 of these patients developed a rash and were advised to cease therapy immediately. One patient cited that the dizziness and blurred vision that he experienced upon starting clonazepam therapy could not be tolerated. 4 patients withdrew later during treatment and identified side effects as an influential factor in their decision to withdraw. These patients cited dizziness, preference for herbal remedies without side effects, impairment of daily work function and feeling cranky and uncoordinated as reasons for cessation. All 4 patients obtained at least partial resolution of their symptoms before withdrawal, with 1 obtaining marked relief, and 1 patient achieving complete resolution of their pain prior to withdrawal.
17 patients actively withdrew from treatment. Reasons for withdrawal cited by patients included side effects (n = 7, discussed above), no relief being obtained (n = 4), pain completely resolved (n = 3), moving interstate (n = 2), and prohibitive costs of travel and consultation fees (n = 1). 16 patients did not attend review appointments within the 6-months preceding June 2009 and were therefore deemed to be lost to follow-up (LTF). All 16 LTF patients had obtained at least partial relief of BMS symptoms during the course of treatment, with 7 achieving marked reduction and 5 obtaining complete resolution of symptoms. 23 patients were still taking clonazepam at their last review appointment within the 6-month period preceding June 2009.
3.6 Blood sample results
14 patients had blood tests taken during the course of treatment. There was no linear relationship observed between the relief obtained and the serum concentration of clonazepam. Levels ranged from below 10 µg/L to 30 µg/L in patients who obtained relief, however two patients with concentrations within these limits did not obtain relief (Table 5).
The cohort of patients included in this study appear to reflect BMS population features documented in existing research in terms of patient age, female gender predominance, main sites affected, onset stimuli, related symptoms, pain severity prior to treatment, and concurrent psychological disease. 1,4,5,8,13
The results of this study demonstrate that dissolving clonazepam tablets orally before swallowing provided marked or complete pain relief in approximately two-thirds of patients who were treated for BMS in this way by two Oral Medicine Specialists from January 2006 to June 2009. While pain scores fluctuated throughout treatment, no patients experienced an increase in pain beyond pre-treatment levels, and more than 80% of patients experienced at least partial resolution of their symptoms over the period in which they were taking clonazepam.
As the mean pain score had reduced by approximately 4 points at the first review appointment, it is evident that distinct pain reduction had occurred for many patients within two months of beginning clonazepam treatment. The 4.5 point mean pain decrease observed over the course of treatment is a significant finding in the context of BMS management, and is higher than changes documented in previous research.4,13
Combining qualitative and quantitative pain change assessments provides a useful means of retrospectively assessing the overall response to treatment within this study group, however an attempt has been made to report the qualitative and quantitative assessment of pain changes separately (Figure 1) to afford full disclosure of their respective contributions to the overall results obtained.
The ‘placebo effect’ for this mode of administration cannot be quantified from the data available, however many potential BMS treatment modalities tested previously have noted only a weak placebo effect, 14,15 with past clonazepam research attributing about 10.7% of the change in pain intensity observed to this factor.4
Side effects associated with the lozenge administration method were mostly transient and mild, however some patients ceased treatment as a result of adverse reactions. Pain resolution, satisfaction with pain reduction, side effects, travel and expenses related to treatment were cited as instigators of therapy cessation among patients who advised the clinics that they were withdrawing from treatment. It is likely that some of these factors may have also contributed to patients being lost to follow up. Since 80% of LTF patients achieved marked reduction of their pain during the course of treatment, positive outcomes are inferred, however it is not possible to identify all the reasons for which clonazepam treatment was ceased (or sought elsewhere) from the available data.
Blood sample results within this study show the effective serum concentration of clonazepam to range from under 10 µg/L to 30 µg/L for reduction of pain associated with BMS, however these findings are confounded by the observation that not all patients with serum concentrations within this range experienced pain relief. While these findings may give a preliminary indication of the serum levels at which clonazepam exerts its pain reducing effect, clearly more research is needed to establish sound serum concentrations limits within which therapeutic effects may be expected. Future dose-concentration-response analysis in this field may benefit from standardization of blood sample timing following medication use, and precise identification of concentrations below the level of 10 µg/L given the frequency of effects in this low range.
Key presenting features including gender, pre-treatment pain intensity and treatment history do not appear to be predictive of the pain resolution that will be achieved via clonazepam lozenge administration. Age may offer some predictive guidance, since younger patients responded more favorably to treatment within this study group. This was a non-uniform yet significant finding, and is a feature that has also been noted in previous research.12
By including patients with perceived, diagnosed and medicated psychological disease, it is intended that this study will afford a more accurate reflection of the population affected by BMS than research that has elected to exclude these groups in the past,4 however it also increases the possibility for medications used for these purposes to interact with the drug of interest in this study. No correlation between patients taking psychoactive or CNS depressant medications and overall pain resolution was observed, however it is possible that drug interactions that were not accounted for may have occurred.
There are obvious and somewhat inherent limitations of retrospective research including reliance on recall, record-keeping and practitioner-dependent standardization that must be considered when interpreting the results of this study. Although this research is unlikely to cover the full spectrum of patient responses to clonazepam administered via a lozenge protocol, this review of the treatment outcomes of 54 patients treated in this way (a relatively large cohort compared with existing BMS treatment research), has documented a clonazepam administration protocol that has not been assessed previously, which appears to be an effective management approach.
A simple comparison with existing research would appear to favor the current protocol compared with systemic or topical administration alone based on pain reduction outcomes. A similar rate of side effects is observed with lozenge administration of clonazepam compared with topical administration alone, however active topical treatment was associated with a slightly lower rate of patient withdrawal due to adverse effects.4 It must be noted that differences in patient numbers, pain assessment criteria, the duration of study and research design may hinder the validity of such comparisons. Whilst the findings of this study are positive, no firm conclusions regarding the efficacy of a lozenge protocol of clonazepam administration relative to alternative modes may be made based on the information currently available.
Given the existing research that now supports the use of clonazepam for the treatment of BMS, and the lack of well-founded alternatives, a quality prospective trial that seeks to compare different modes and establish guidelines for appropriate administration (in terms of dosage range and serum thresholds) seems warranted. There are several theories that attempt to explain the means by which clonazepam exerts its pain reduction effects, yet significant uncertainty still exists.4,12,14 Research that seeks to directly compare different modes of clonazepam delivery may help to reveal the mechanisms of drug action, and perhaps also improve our understanding of the etiology of this condition. Considering that the benzodiazepine-GABAA receptor complex may be a strong candidate to explain the pain reducing activity of clonazepam, location of these receptors in both central and peripheral tissues may provide a tentative foundation for the further exploration of a combined topical and systemic delivery mode.4,15
Multi-centre co-operation and establishment of benchmarks for pain reduction assessment will afford increased opportunities for collaboration and comparison of findings to accelerate knowledge in the field of BMS management. While most existing research relies predominantly on basic pain assessment scales as an indicator of treatment success, our current understanding of the psychosocial influence of pain would view such an analysis as incomplete. For this reason, future research may benefit from consideration of the impact of pain on quality of life, for example, one’s ability to work, engage in relationships, complete daily tasks, and establish healthy coping mechanisms for residual symptoms. Interpretation of success in this light may help to identify holistically sound treatment approaches to address the chronic and often debilitating pain associated with BMS.1 The method of dissolving clonazepam tablets orally before swallowing appears to be a promising management option based on the preliminary findings outlined in this paper, and is certainly a strong candidate for further investigation.
Quantitative and qualitative assessment of pain reduction
|Reduction||Quantitative||Qualitative terms used|
|None||0%||Not helping; no improvement.|
|Partial||<50%||Partial resolution; progressing well; some improvement.|
|Marked||≥50%||Marked improvement; mostly symptom free; most symptoms resolved; treatment successful; not tingling or burning most of the time.|
|Complete||100%||Complete resolution; pain free; no more pain.|
Daily doses used
|Dose (mg/day)||Frequency||Percentage of total prescriptions (n = 169)|
Summary of pain scores recorded before treatment, at the first review appointment and at the final appointment.
|Pain scores||Mean||SEM||Median||Range||Significant change *|
(n = 42)
6.770.2573 to 10N/A1st review
(n = 39)
2.830.3130 to 6YesFinal review
(n = 36)
2.070.3120 to 7Yes
* Mann-Whitney test (P < 0.05)
Presenting data for different response groups
(n = 6)Partial reduction
(n = 9)Marked reduction
(n = 21)Complete resolution
(n = 15)Side effects
(n = 3)P-valueAge52.1765.0058.0051.0065.33Sig. *Females/ Males6/07/215/613/23/0NSDuration of symptoms34.6712.138.7820.2713.50NSPre-treatment
pain intensity6.256.166.927.007.00NSPractitioners involved before presentation2.002.001.381.602.00NSNumber of concomitant
* Patients with complete resolution were significantly younger than patients with only partial pain reduction (two-tailed Mann-Whitney test, P = 0.007).
Blood sample results
Serum concentration of clonazepamTime since last dose
19 µg/L – <20 µg/L
5 µg/L – 11µg/L
<10 µg/L – 30 µg/L
<10 µg/L – 12 µg/L
Summary of pain resolution observed.
* Patients withdrew within 10 days of starting clonazepam therapy, before assessment of pain changes.
1. Zakrzewska JM, Forssell H, Glenny AM. Interventions for the treatment of burning mouth syndrome. Cochrane Database Syst Rev 2005;CD002779.
2. Lipton JA, Ship JA, Larach-Robinson D. Estimated prevalence and distribution of reported orofacial pain in the United States. J Am Dent Assoc 1993;124:115-121.
3. Tammiala-Salonen T, Hiidenkari T, Parvinen T. Burning mouth in a Finnish adult population. Community Dent Oral Epidemiol 1993;21:67-71.
4. Gremeau-Richard C, Woda A, Navez ML, Attal N, Bouhassira D, Gagnieu MC, et al. Topical clonazepam in stomatodynia: a randomised placebo-controlled study. Pain 2004;108:51-57.
5. Gorsky M, Silverman S. Jr, Chinn H. Clinical characteristics and management outcome in the burning mouth syndrome: an open study of 130 patients. Oral Surg Oral Med Oral Pathol 1991;72:192-195.
6. Zakrzewska JM. The burning mouth syndrome remains an enigma. Pain 1995;62:253-257.
7. Basker R, Sturdee D, Davenport J. Patients with burning mouths. A clinical investigation of causative factors, including the climacteric and diabetes. Br Dent J 1978;145:9-16.
8. Bergdahl M, Bergdahl J. Burning mouth syndrome: prevalence and associated factors. J Oral Pathol Med 1999;28:350-354.
9. MIMS online prescribing information. URL: ‘http://mims.hcn.net.au.ezproxy.library.uq.edu.au/ifmx-nsapi/mims-data/?MIval=2MIMS_abbr_pi&product_code=3493&product_name=Paxam’. Accessed 16/06/08.
10. Harkins S, Linford J, Cohen J, Kramer T, Cueva L. Administration of clonazepam in the treatment of TMD and associated myofascial pain: a double-blind pilot study. J Craniomandib Disord 1991;5:179-186.
11. Grushka M, Sessle BJ. Burning Mouth Syndrome. Dent Clin North Am 1991;35:171-184.
12. Grushka M, Epstein J, Mott A. An open-label, dose escalation pilot study of the effect of clonazepam in burning mouth syndrome. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;86:557-561.
13. Woda A, Navez ML, Picard P, Gremeau C, Pichard-Leandri E. A possible therapeutic solution for stomatodynia (burning mouth syndrome). J Orofac Pain 1998;12:272-278.
14. Forssell H, Jaaskelainen S, Tenovuo O, Hinkka S. Sensory dysfunction in burning mouth syndrome. Pain 2002;99:41-47.
15. Drugan R, Holmes P. Central and peripheral benzodiazepine receptors: involvement in an organism’s response to physical and psychological stress. Neurosci Biobehav Rev 1991;15:277-298.