Accepted Manuscript World Workshop on Oral Medicine VI: Clinical implications of medication-induced salivary gland dysfunction Ardita Aliko, DDS, PhD, Associate Professor, Andy Wolff, DMD, Colin Dawes, BDS, PhD, Professor, Doron Aframian, DDS, PhD, Professor, Gordon Proctor, BSc, PhD, Professor, Jörgen Ekström, MD, PhD, Professor, Nagamani Narayana, DMD, MS, Associate Professor, Alessandro Villa, DDS, PhD, MPH, Assistant Professor, Ying Wai Sia, DMD, Revan Kumar Joshi, Richard McGowan, Siri Beier Jensen, DDS, PhD, Associate Professor, A. Ross Kerr, DDS, MSD, Clinical Professor, Anne Marie Lynge Pedersen, DDS, PhD, Associate Professor, Arjan Vissink, MD, DMD, PhD, Professor PII:

S2212-4403(15)00569-6

DOI:

10.1016/j.oooo.2014.10.027

Reference:

OOOO 1156

To appear in:

Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology

Received Date: 24 September 2014 Accepted Date: 12 October 2014

Please cite this article as: Aliko A, Wolff A, Dawes C, Aframian D, Proctor G, Ekström J, Narayana N, Villa A, Wai Sia Y, Kumar Joshi R, McGowan R, Beier Jensen S, Kerr AR, Pedersen AML, Vissink A, World Workshop on Oral Medicine VI: Clinical implications of medication-induced salivary gland dysfunction, Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology (2015), doi: 10.1016/ j.oooo.2014.10.027. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT World Workshop on Oral Medicine VI: Clinical implications of medication-induced salivary gland dysfunction

Ardita Aliko, Associate Professor, DDS, PhDa,b, Andy Wolff, DMDc, Colin Dawes,

RI PT

Professor, BDS, PhDd, Doron Aframian, Professor, DDS, PhDe, Gordon Proctor, Professor, BSc, PhDf, Jörgen Ekström, Professor, MD, PhDg, Nagamani Narayana, Associate Professor, DMD, MSh, Alessandro Villa, Assistant Professor, DDS, PhD, MPHi, Ying Wai Sia, DMDj, Revan Kumar Joshik, Richard McGowanl, Siri Beier Jensen,

SC

Associate Professor, DDS, PhDm, A. Ross Kerr, Clinical Professor, DDS, MSDl, Anne Marie Lynge Pedersen, Associate Professor, DDS, PhDm, Arjan Vissink, Professor, MD, DMD,

M AN U

PhDn.

Affiliations a

Faculty of Dental Medicine, University of Medicine, Tirana, Albania; bBroegelmann Research

TE D

Laboratory, Department of Clinical Science, University of Bergen, Bergen, Norway; cTel-Aviv Sourasky Medical Center and Saliwell Ltd., Israel; dDepartment of Oral Biology, University of Manitoba, Canada; eThe Hebrew University, Israel; fDental Institute, King's College London, g

Department of Pharmacology, The Sahlgrenska Academy at the University of

EP

UK;

Gothenburg, Sweden; hDepartment of Oral Biology, UNMC College of Dentistry, USA; Department of General Dentistry, Henry M. Goldman School of Dental Medicine, Boston

AC C

i

University, Boston, USA; jMcGill University, Canada; kDAPMRV Dental College, Bangalore, India; lNew York University College of Dentistry, USA; mDepartment of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark; nUniversity of Groningen and University Medical Center Groningen, The Netherlands Correspondence to: Dr. Andy Wolff, Tel-Aviv Sourasky Medical Center and Saliwell Ltd., 65 Hatamar St., Harutzim 60917, Israel; Email: [email protected]; Phone: +972-50-880-1852 Disclosures: No conflicts of interest. 1

ACCEPTED MANUSCRIPT Word count, abstract: 199 Word count, text: 7444 Number of reference: 172 Number of tables: 6

RI PT

Number of figures: 1 Abstract

Objective. This study aimed to systematically review the available literature on the clinical implications of medication-induced salivary gland dysfunction (MISGD).

SC

Study design. The systematic review was performed using PubMed, Embase and Web of Science (through June 2013). Studies were assessed for degree of relevance and strength of

M AN U

evidence, based on whether clinical implications of MISGD were the primary study outcomes, as well as on the appropriateness of study design and sample size. Results. For most purported xerogenic medications, xerostomia was the most frequent adverse effect. In the majority of the 129 reviewed papers, it was not documented whether

TE D

xerostomia was accompanied by decreased salivary flow. Incidence and prevalence of medication-induced xerostomia varied widely and was often associated with number and dose of medications. Xerostomia was most frequently reported to be mild-to-moderate in

EP

severity. Its onset occurred usually in the first weeks of treatment. There was selected evidence that medication-induced xerostomia occurs more frequently in women and elderly,

AC C

and that MISGD may be associated with other clinical implications such as caries or oral mucosal alterations.

Conclusions. The systematic review showed that MISGD constitutes a significant burden in many patients and may be associated with important negative implications for oral health.

Statement of clinical relevance This article describes a systematic review addressing the clinical implications of medicationinduced salivary gland dysfunction.

2

RI PT

ACCEPTED MANUSCRIPT

Introduction

Saliva is of paramount importance for the maintenance of oral health. The complex and unique mixture of water, proteins, glycoproteins and ions aids in several functions including

SC

lubrication of oral tissues, wound repair, oral acid neutralization and maintenance, enamel maturation in children and tooth remineralization. The antimicrobial activity of saliva is

M AN U

generally linked to the mechanical clearance of food and bacteria, acting as a physical barrier, and as a result of the vast array of its immune and non-immune defense components1. Saliva also exerts digestive functions, including dissolution of taste compounds, trophic functions in the maintenance of taste buds, facilitation of mastication and

TE D

swallowing, initial digestion of starches and lipids and formation of a food bolus2. Studies provide evidence that the functions of saliva are not restricted just to the mouth. It has been proposed that saliva helps in esophageal clearance and gastric acid buffering3. Further, it

EP

facilitates speech and is an important mediator in social interactions. Salivary gland dysfunction resulting in reduced salivary flow rate (salivary

AC C

hypofunction) and/or compositional changes often has significant consequences, including xerostomia (i.e., the subjective feeling of dry mouth) and other sensory disturbances, an increased risk of dental caries and candidiasis, and difficulties in speech, mastication, swallowing and denture retention4, followed by substantial impact on nutritional status5. The chronic course of most cases of salivary gland dysfunction and related complaints contributes to their pervasive and negative effects on the quality of life, independent of comorbidity6. A wide range of systemic disorders can directly or indirectly result in salivary gland dysfunction, such as Sjögren’s syndrome, diabetes, depression, head and neck radiotherapy 3

ACCEPTED MANUSCRIPT and acquired immune deficiency syndrome. However, the most commonly recognized cause for salivary gland dysfunction is the use of medications7. Further, there is general agreement that intake of medications is a better predictor of risk status for xerostomia than either age or gender8,9.

RI PT

Salivary glands are supplied with a number of various types of receptors, including acetylcholine muscarinic receptors, salivary glands contain target sites for many medications10. Indeed, more than one thousand drugs are reported to be associated with xerostomia as an adverse effect7, and are denoted as “xerogenic” in the medical literature.

SC

Tricyclic antidepressants, muscarinic receptor antagonists, antipsychotics, opioids and benzodiazepines, antihypertensives and antihistamines are the most commonly implicated10.

M AN U

Despite being widely acknowledged as a significant clinical problem, the spectrum of symptomatology and clinical oral implications related to medication-induced salivary gland dysfunction (MISGD) needs further characterization. Because of their heterogeneous nature and possible interactions, multiple mechanisms may exist whereby medications can

TE D

influence the quantity and quality of saliva. Thus MISGD may give rise to different patterns and severity of oral complications, as compared with the more predictable and better documented cases of Sjögren’s syndrome and radiation-induced salivary gland dysfunction.

EP

Therefore, the present paper systematically reviews current literature on clinical implications

AC C

of MISGD in the oral region.

MATERIALS AND METHODS The group on MISGD within the World Workshop on Oral Medicine 6 (WWOM VI) is comprised of 5 reviewers (AA, RJ, NN, YS and AlV), 6 consultants (senior experts in fields related to MISGD: DA, CD, JE, AMP, GP and ArV), one research librarian (RM), a Group Head (AW) and 2 supervisors on behalf of the WWOM VI Steering Committee (SBJ and ARK). This, review addresses one of the MISGD topics covered by the group, the clinical implications of MISGD, except drooling/hypersalivation (which be covered separately). The research method was based on the policies and standards set forth by a Task Force for 4

ACCEPTED MANUSCRIPT WWOM IV11 and by the PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses)12, adapted to the current systematic review. Step 1 - Scope definition: The current systematic review covered 2 research questions: 1) What is the spectrum of the symptomatology that is related to MISGD?

RI PT

2) What are the sequelae that affect oral health in patients with MISGD? Step 2 – Keyword selection: Keywords were selected for each research question.

For question 1, keywords were: medication/drugs AND salivary glands/saliva AND dry mouth, xerostomia, hyposalivation, burning mouth, taste, eating, swallowing, speaking,

SC

nutrition, quality of life, depression, withdrawal, social isolation, hoarseness, lip cracking,

candidiasis, oral infections.

M AN U

fissured tongue, dysphagia, stomatodynia, dysgeusia, oral burning, dental caries, candidosis,

For question 2, keywords were: medication/drugs AND salivary glands/saliva/dry mouth/xerostomia/hyposalivation AND caries, decayed teeth, oral candidiasis, oral candidosis, oral infection, salivary gland enlargement, tender oral mucosa, dentures,

TE D

halitosis, oral fetor, oral malodor, periodontitis, oral ecology, oral microflora, lobulated tongue, tongue atrophy, dental erosion, taste, angular cheilitis, bald tongue, fissured tongue, fungal infections, upper gastrointestinal tract.

EP

Step 3 – Literature search: Our literature search was conducted, through June 2013, in the PubMed, Embase and Web of Science databases, based on our chosen keywords and

AC C

subject headings, and was not limited in time, publication type or language. Group members were also encouraged to submit articles of interest located from referral or hand-searching. The search was completed by a hand search of the reference lists in the eligible papers, and after removal of duplicates, a total of 747 records was retained for Step 4. Step 4 – Record screening for eligibility: Each of the 747 records was screened independently by four reviewers (AA, NN, AlV and YS), supervised by two consultants (ArV for "Symptoms", AMP for "Oral sequelae"), and retained for further analysis or excluded based on their relevance to any of the research questions. 570 papers were found to be relevant to both topics. 5

ACCEPTED MANUSCRIPT Step 5 – Paper selection for type of study, relevance and level of evidence: This step started with calibration among the reviewers in order to ensure that they applied similar standards in the performance of their reviews. Papers were then divided among the reviewers, who analysed publication titles, abstracts and the materials and methods sections

RI PT

for key parameters: (a) the degree of relevance, (based on whether xerostomia or other clinical implications of MISGD were the primary outcomes of the study or not), (b) the strength of methodology provided in the paper (according to the appropriateness of the study design and sample size) and (c) the level of precision by which the medications were

SC

classified, ranging from organ or system treated (lowest level), through indication (middle level) to chemical substance (highest level). The highest level of evidence was intended for

M AN U

papers studying MISGD as primary outcome, reporting meta-analyses, systematic reviews or randomized clinical trials, and focusing on chemical substances. As a result of this step, 365 papers were retained for full-text analysis. The remaining papers were excluded because of poor relevance and/or weak methodology.

TE D

Step 6– In-depth analysis: This was based on expert interpretation of the evidence. Supervised by the consultants and by the Group Head, reviewer AA screened the remaining 365 selected publications on reading the full-text. Another 236 papers were excluded for

EP

such reasons as reporting case studies or assessing MISGD as minor outcome in importance. This resulted in a total of 129 papers, which were analyzed and used for

AC C

preparation of the current systematic review (Figure 1).

RESULTS

Results of the search strategy and critical appraisal of the included studies On the basis of our assessment system, none of the studies had the highest level of evidence and all of them displayed a considerable risk of bias. As the risk of bias was caused either by poor relevance (MISGD was not the primary outcome) and/or weak methodology (studies other than meta-analyses, systematic reviews or randomized controlled trials) and/or 6

ACCEPTED MANUSCRIPT low precision level of the medication classification (no classification by chemical substance), it was unmeasurable. Therefore the retained papers were not classified by level of evidence. Of the 129 eligible papers, 122 reported data on xerostomia, three studies reported data on other MISGD-related complaints, and seven studies contained information on oral sequelae

RI PT

of MISGD. Xerostomia was systematically assessed in only 16 studies and the methodology for recording xerostomia was quite variable. In some of the studies a single question had been used; either “Does your mouth usually feel dry?” or “Do you have dry mouth”13-19. Questionnaires regarding dry-mouth-related subjective symptoms and behaviors assessed

SC

either by a categorical scale20-25 or a visual analogue scale (VAS)21,26-28 have also been used. However, most of the studies included in this review were trials, testing the safety and

M AN U

efficacy of certain medications for treating a variety of diseases, either randomized or openlabel. In these studies, salivary gland dysfunction or xerostomia was not the primary outcome and, moreover, it is uncertain whether it was systematically assessed as a secondary outcome. Most probably, the inclusion of yes/no questions of “dry mouth” were just based on

TE D

listing of many symptoms potentially associated with the test medications. Furthermore, in the majority of papers, it was not mentioned whether xerostomia was accompanied by an actual decrease in salivary flow rate.

EP

Studies evaluating the relation between MISGD and dental caries or other oral sequelae also varied in their methodology. Caries status was evaluated by either measuring

AC C

the root caries prevalence29, the incidence of caries and decayed and filled tooth surfaces score30, or the root caries index31,32. Oral mucosal status was evaluated in three studies, either by assessment of palatal inflammation33, oral mucosal inflammation score34, or by recording the presence/absence of oral mucosal lesions of any type22.

Xerostomia Incidence/prevalence of xerostomia A total of 85 studies included in the review reported incidence or prevalence data for medication-induced

xerostomia17,23-26,35-114.

Incidence

data

were

available

from

56 7

ACCEPTED MANUSCRIPT randomized trials (RCTs), five pooled analyses of RCTs and 17 open-label or prospective studies. The incidence ranged from 0.14 to 73.8% in the treatment group. It was generally lower in the placebo group, ranging from 0 to 31.4%. Prevalence data were available from seven cross-sectional studies, of which two were population-based. The prevalence of

RI PT

xerostomia given as the percentage of medicated persons, regardless of the type of medication used, was 32.1%17 and 27.4%24 for these two population-based studies. The remaining cross-sectional studies reported prevalence data for patients on specific medication categories and varied from 7.5% among patients taking cardiovascular

SC

medications65 to 35.1% in patients on highly active antiretroviral therapy (HAART)70, 50% in patients taking antihypertensives94 and up to 71% in patients taking different types of

M AN U

antidepressants72. Xerostomia was also a quite common symptom in the non-medicated population, as it was found in 16.9%17, 14.4%24 and in 25.5% of persons94, although the difference from medicated persons was statistically significant. The difference in xerostomia prevalence between medicated and non-medicated persons was more pronounced in a

TE D

younger population of age 32 years, being 25.2% and 5.0%, respectively23.

Time to onset of xerostomia

EP

None of the included studies reported an increase in the incidence of xerostomia with longterm treatment (Table 1). On the contrary, incidence of xerostomia was higher during the

AC C

initial phase of the treatment, but notably lower with prolonged treatment in most studies36,83,108,115-118. In two of the studies which combined a double-blind period followed by an open-label period, the incidence of xerostomia during the open-label period was found to be lower in persons who had already received active treatment during the double-blind period than in those initially randomized to placebo, and did not increase with prolonged treatment119,120. By contrast, the hierarchical linear modelling used in the study of Mavissakalian et al. showed similar values for the predicted mean probabilities of imipramine-induced xerostomia at the end of the initial open-label period (0.68) as compared with the end of the following randomized double-blind period (0.60)121. 8

ACCEPTED MANUSCRIPT

Relation of xerostomia to medication dose Characteristics of studies included in this review and collective data regarding relation of xerostomia and medication dose are summarized in Table 2. The incidence of xerostomia a

dose-related

increase

across

10

investigations42,102,104.

Relation of xerostomia to the number of medications

studies76,88,122-129,

but

not

in

three

RI PT

showed

SC

Data on the relation of xerostomia to the number of medications taken were available from 15 studies (Tables 3 and 4). In three of the four randomized studies included, the incidence

M AN U

of xerostomia in patients receiving two medications was higher compared with patients receiving one medication plus placebo122,130,131, whereas in another study it was lower, but that difference was not statistically significant132 (Table 3). The remaining studies were epidemiological (Table 4). Polypharmacy (use of more than one medication) was significantly

TE D

associated with high incidence or prevalence of xerostomia in all but one study14. The latter study differed significantly in that the patients’ ages were notably higher (mean age 84.6±8.4 years), compared with the other studies. Additionally, the method for recording xerostomia in

EP

this study was less sensitive, since xerostomia was considered present only if the patient had

AC C

both a dry mouth and a frequent feeling of thirst, which often are non-related phenomena.

Severity of xerostomia

Characteristics of the studies that reported data on the severity of medication-induced xerostomia are summarized in Table 5. In most studies the severity of xerostomia was evaluated by asking patients to rate their feeling of dry mouth as mild, moderate, severe, or either intolerable134, or excessive as noted by parents and caregivers in a study of children with developmental disabilities135. In two other studies, xerostomia severity was measured by a VAS score26 and by a 0-10 scale25. Most xerostomia reports in these collective studies were mild to moderate in intensity, with the exception of two investigations, in which 9

ACCEPTED MANUSCRIPT xerostomia was found to be excessive134, or intolerable135. Medication discontinuation rates due to xerostomia were available from 11 studies and ranged from 0 to 4.9% in RCTs and from 0.02% to 12% in non-RCTs (Table 5).

RI PT

Clinical course of xerostomia Data on the clinical course of medication-induced xerostomia were available from four studies. Barton et al.139, demonstrated that prevalence of xerostomia remained fairly stable throughout the use of venlafaxine (for treatment of hot flashes) with low dosing regimens; by

SC

contrast, the group that used 150 mg/day had a peak prevalence of xerostomia at the end of the 4 week randomized period which gradually diminished during the open-label period to

M AN U

levels similar to those in the other groups, which suggests that some of the existing cases of xerostomia in the 150 mg/day group resolved. Feltner et al.116 also found that the cases of xerostomia as an adverse effect from taking pregabalin actually remitted within the first few weeks, and xerostomia had a median duration of 38 days. In a small 9 week open-label

TE D

study of bupropion involving 26 patients, it was found that, of the 7 cases with medicationinduced xerostomia, all resolved within one week with the exception of 3 cases with persistent xerostomia64. In contrast, it has been demonstrated that xerostomia severity, as

EP

evaluated by VAS score at baseline, at 4 weeks and at 10 weeks of treatment, worsened during the course of a double-blind trial in groups treated with either oxybutynin or

AC C

tolterodine, p1 second generation antipsychotic: 84/162 (52%) 1 second generation antipsychotic: 503/1796 (28%) p