The Efficacy of Potassium Salts As Agents for Treating Dentin Hypersensitivity

TOPICAL RF.VIFW The Efficacy of Potassium Salts as Agents for Treating Dentin Hypersensitivity

Robert Orchardson, BDS. PhD Formulations containing potassium salts (eg, chloride, nitrate, cit- Senior Lecturer rate, oxaiate) are widely used for treating dentin hypersensitivity Division of Neuroscience and (DH). The purpose of this review was to evaluate evidence for the Biomédical Systems cUnical efficacy of potassium salts in reducing DH and also to Institute of Biomédical and Life consider the biologic basis for any effects. Literature searches were Sciences used to identify reports of clinical trials of potassium-containing University of Glasgow Glasgow, Scotland, United Kingdom preparations. Searches revealed 3 trials of potassium nitrate solutions or gels; 2 trials of mouthivashes containing potassium nitrate David G. Gillam. BDS. MSc, DDS or citrate; 6 trials of potassium oxalates; and 16 double-blind ran- Senior Research Fellovu and Lecturer in doinized trials of toothpastes containing potassium nitrate, chlo- Peri odontology ride, or citrate. The toothpaste studies provided quantitative data Department of Penodontology on treatment effects. These outcome measures were expressed as Eastman Dental Institute for Oral Health percentage reductions in sensitivity to cold air and mechanical Care Sciences stimulation and the patients' subjective reports. Trials of topically London, England, United Kingdom applied solutions yielded inconsistent results. Potassium-containing mouthivashes produced significant reductions in sensitivity. All Correspondence to; potassium-containing toothpastes produced a significant reduction Dr R. Orchardson in sensitivity to tactile and air stimuli, as well as subjectively Laboratory of Human Anatomy University of Glasgow reported sensitivity, ¡n most studies, the active agent (potassium) Glasgow, Scotland was superior to the minus-active control (placebo), but a few of G12 8OQ United Kingdom the more recent trials have demonstrated significant placebo E-mail: [email protected] effects. It ¡s postulated that potassium ions released from toothpastes diffuse along the dentinal tubules to inactivate intradental nerves. However, this principle has never been confirmed in intact human teeth. The mechanism of the desensitizing effects of potassium-containing toothpastes remains uncertain at present. J OROFAC PAIN 2000;14:9-I9.

Key words: dentin hypersensitivity, desensitization, intradental nerves, potassium, toothpastes

entm hypersensitivity (DH) is characterized by short, sharp pains that arise typically when thermal, evaporative, Dmechanical, or osmotic stimuli are applied to exposed dentm, and which cannot be explained by any other form of dental defect ot pathology.' The majotity of "natural" stimuli that cause pain in hypersensitive teeth appear to excite lntradental nerves indirectly by increasing fluid flow in dentinal tubules.--^ Exceptions to this principle are electric current and thermal stimuli, which can directly activate the s ma Her-díame ter nerve axons in the pulp.''-' This hydrodynamic principle prompts 2 basic approaches for treating hypersensitive dentin: (1) occlude patent tubules to reduce any stimulus-evoked fluid flow, or (2) reduce lntradental nerve excitability so that the nerves do not respond to the stimulus-evoked fluid movements. A wide range of agents and procedures have been used to treat dentin hypersensitivity,'^'^ but none are completely effective in every instance. The aims of this

Journal of Orofacial Pain 9 Otchardson/Gillam review ¡ire (1) to evaluate the clinical evidence that a compound mouthrinse containing 2% potassium potassiutn salts are effective dentlnal desensitizing citrate, 0.05% cetylpyridinium chloride, and agents and (2) to consider the possihle biologic 0.05% NaF with an identical base formulation basis tot these effects. lacking these active ingredients. These investigators found that both mouthrinses produced highly significant reductions in tooth sensitivity to tactile Review Procedure and airblast stimulation. However, they failed to detect any significant differences between the We searched the literature to identify reports of desensitizing effectiveness of the test and minus- clinical trials of potassium-containing prepara- active or placebo formulations. This latter study is tions, including toothpastes, mouthwashes, and of interest, as it confirmed the strong "placebo topical Lipplications. This lnfotmation has been tesponse" evident in other recently published clinical trials of desensitizing toothpastes.^^'^"^ augmented by material from otit own reference libraries. Although most of the information surveyed was taken from full papers in peer-reviewed Potassium Oxalates publications, unrefereed abstracts were also considered if they provided sufficient information Greenhill and Pashley" investigated rhe effects of about methods and results. The review also covers a variety of desensitizing agents on the hydraulic labotatory-based material that is relevant to conductance (permeability) of dentin in vitro. understanding the possible biologic mechanisms of These investigators found that out of 29 agents the varions clinical agents. tested, 30% potassium oxalate solution caused the gteatest reduction (98%) in dentin permeability. Since then, the basic dentin disc pteparation has Potassium as a Desensitizing Agent been used widely to assess the tubule-blocking ability of potential desensitizing agents in vitro.'^ Solutions, Gels, and Mouthrinses On the basis of its effectiveness in reducing dentin permeability, potassium oxalates have been used Potassium salts such as caustic potash (potassium for treating DH. The rationale for the method is hydroxide) were used over 100 years ago as obtun- that soluble potassium oxalate is converted to dents ptior to the introduction of local insoluble calcium oxalate within dentinal anesthetics.^ However, these agents fell out of tubules.'^ favor for nearly 100 years. In a largely anecdotal In clinical trials, dipotassium oxalate (Kj account, Hodosh^ stated that topical applications oxalate'**), monopotassium monohydrogen oxalate of 1 to 15% potassium nitrate (KNO^], saturated (KH oxalate-^**), and a combination of the 1^' have solutions of KNO^, ot a paste containing 10% been reported to cause significantly greater reduc- KNOj were effective m reducing den tin hypersen- tion in sensitivity than minus-active controls. sitivity. In a single-blind study, Green et al"* found Gillam et aP^ found that K, oxalate (Protect, J. 0. that topical applications of a calcium hydroxide Butler) and All Bond 2 resin (Bisco) both teduced paste or a KNO. paste (concentrations not speci- DH, but their effects were not significantly differ- fied) gave significantly greater reductions in sensi- ent. Gooley and Sandoval'' and Cuenin et aF'* tivity to mechanical and hot stimuli compared evaluated the desensitizing efficacy of 3% KH with the control (sahne). Calcium hydroxide was oxalate solutions. These investigators reported that more effective in reducing pain on cold stimula- their respective control agents of distilled water tion. Reinhart et al" found that 4 weeks of self- and 3% sodium chloride (at pH 2.4) were more administered daily applications of a 10% KNO,- effective in reducing sensitivity than KH oxalate. glycerin gel were less effective in reducing dentin Thus, although in vitro studies suggest that potas- sensitivity than was a plain glycerin gel. sium oxalate would be an ideal desensitizing agent, There is limited information about the efficacy clinical studies have tended to be inconclusive.-^ of potassium-containing desensitizing mouthwashes.•"- Gillam et aP^ reported that a 3% Toothpastes KNOj/silica/sodium fluoride (NaF) mouthrinse caused a significantly greater reduction in DH to For the purpose of this review, 16 full papers were mechanical and thermal stimuli than did a identified that reported outcomes of clinical trials silica/NaF control mouthwash. More recently, on toothpastes containing potassium salts (5% Yates et aP'' compared the desensitizing efficacy of KNO,, 3.75% potassium chloride [KCI], or 5.3%

10 Volume 14, Number 1, 2000 Orchardson/Gillam potassinm citrate). The details of the studies are was as effective as a standard 5% KNO3 paste in summarized in Table 1. All studies used doulile- reducing DH. Schiff et aP' have reported that a blind, randomized designs. The studies also dentifrice containing 5% KNO^ and 1,500 ppm employed similar exclusion criteria. There were Na MFP in a precipitated calcium carbonate base some differences among studies in the inclusion was more effective than a control in reducing DH. criteria, such as nnmbers and types of sensitive Salvato et aP" reported that a paste containing teeth, the numbers of subjects per group, and the 3.75% KCl and 0.8% Na MFP was significantly duration of the trial, as well as the modes of sensi- more effective than a control paste lacking botb tivity measurement and the assessment intervals KCl and Na MFP. To counter claims that the ¡Table I). Most included minus-active (placebo) desensitizing effects could have been due to the Na controls ¡efficacy studies), although a few included MFP in the paste, Silverman et al'^ tested 2 pastes only positive controls (equivalence studies). These containing 3.75% KCl. One of these contained differences in the trial design made it difficult to 0.8% Na MFP. Both of the KCI-containing pastes attempt quantitative comparisons. However,, when vvere shown to be significantly more effective than analyzed, these studies provided sufficient raw a control paste that contained all the ingredients of data to enable average percentage changes in sensi- the other pastes, except KCl and Na MFP. Gillam tivity CO be calculated for the active and control et al''' compared a KCI/MFP toothpaste and a formulations. The analysis is confined to the most strontium acetate/NaF toothpaste with a control commonly nsed and generally accepted sensitivity paste containing Na MFP, calcium glycerophos- measures-': tactile stimulation ¡Tables 2 and 3), phate, and NaF. All 3 pastes produced a signifi- air-jet stimulation (Table 4), and global subjective cant reduction in DH compared to baseline; how- evaluations (Table 5). ever, the investigators were unable to demonstrate Several studies have reported that a 5% KNO, any significant differences between tbe 3 formula- paste was significantly more effective In reducmg tions. sensitivity' than a placebo or mmus-active control All studies surveyed have reported that toothpaste.-*"^- Tarbet et aP^ reported that a 5% KNO, pastes containing 5% KNO3 or 3.75% KCl signifi- paste was superior to 3 other desensitizing pastes cantly reduce DH to tactile (Tables 2 and 3) and containing 10% strontium chloride ¡SrCl.,), 2% air-jet (Table 4) stimuli as well as sub|ective mea- sodium citrate, or 1.4% formaldehyde. Mano- sures (Table 5). In most studies, these pastes are chehr-Pour et al,-'"* Chesters et al," and West et significantly better than the minus-active or aP^ reported that whereas pastes containing 5% placebo controls. However, in 4 studies,^^•'^•^'*'^' KNOj caused a significant reduction in DH, the both the potassium pastes and the control pastes performance of this agent was not significantly dif- produced a significant reduction in sensitivity, but ferent from that of the controls ("minus-active" there was no significant difference between the formulation or a paste containing sodium monoflu- performance of the KNO3 or KCl pastes and the orophosphate [Na MFP]). Some other inconsisten- control pastes. Why should the results of these tri- cies were identified. Silverman et aF- and Tarbet et als be different from the results of the other trials aP^ reported that a 5% KNO3 paste was superior analyzed? Differences in trial outcome could to a 10% SrClj paste. However, in a 12-week reflect variations in the trial designs.'"' However, semi-quantitative study, Collins et al-^ reported the design parameters of these 4 trials'^'^^'^*''^ lay that pastes containmg 5% KNÜ, and 10% SrCI, within tbe spread of values evident in all the trials were comparable in their desensitizing effective- analyzed (Table 1). Nor are the differences likely ness. Person et aP^ also showed tbat toothpastes to be due to variations ¡n the initial sensitivity lev- containing 5% KNO, + 0.76% Na MFP and 10% els, as in all studies analysed there were no signifi- SrCI, caused a significant increase in the sensitivity cant differences between the baseline sensitivities thresholds to warm and hot air stimuii after 4 and of the potassium paste and control groups. It is 8 weeks of use, but the performances of these 2 possible that differences in trial duration could pastes were not significantly different. account for variations in the results.""* The median Schiff et al-'" evaluated a combination paste con- duration of the 16 trials analyzed was 8 weeks. taining 5% KNO3, 1.3% soluble pyrophosphatc, The studies by Gillam et al'^ and West et al'* 1.5% copoiymer of methoxycellulose and maleic lasted 6 weeks, while the trials of Chesters et al" acid, and 0.243% NaF. This paste was signifi- and Manochehr-Pour et al^** lasted 8 and 12 cantly more effective than the control paste, which weeks, respectively. The trial by Chesters et aP^ contained all ingredients except the 5% KNO,. was the only one to include K cittate as one of the Ayad et aP" reported that this combination paste test agents. This trial also employed a novel form

Joumal of Orofacial Pain 11 Orchardson/Gillam

s; -g

5 g o ó til < > o-gë SJG.^S Eíás w¿ 0)5 ^ S — 5'^s

^ Sí -S-

y ó

P. S — LTî m 00-s

O E 5,0

il sí i ^to b m t c .E Q..E Q.-E D.-? Q.:

H .-. 51 ^ 01 (Ü .^ i]> 'P OJ

X a- /\l S- Al 3- Al 3- Al S- X 3- X S- Al a- Al S- Al

PJ IN

Ol O (D CM en

w CO ^ n i-i d ' = = a ' : g Ë 'r II . o ÏÏ Sin o> ai TU í <

iô Ö 5 (ÍÍ Ü ^ .!• I i

12 Volume 14, Number 1,2000 Orchardson/Gillam

Table 2 Percentage Reduction in Sensitivity to Tactile Stimulus for Potassium- Containing Dentifrices and Controls ¡Placebo Pastes), Measured by Intensity Scale Active; Placebo; reduction reduction Active > placebo Snidy Active agent* from baseline from baseline {P value)

Manochehr-Pour 5% KNO^tl) 43% 8% P=0.3 etal, 1984''' 5% KNO312) 47% P=0.21 Nagata et al, 1994^^ 5% KNOa ") 82% 22% P<0.01 Gillametal. 1996' = 3.75% KCI <3) 48% 46% P=0.49 Westetal, 1997'^ 5% KNO, (6) 57% 52% P = 0.85 •1 = Danquel, 2 = Sensodyne F/Promise; 3 = Sensodyre F, 4 = Sensilive/Tartar Conlrol, 5 = Sensiliviiy Prote Crest; 6 - Aquafresh Sensitive; 7 = brand not specified.

Table 3 Percentage Reduction in Sensitivity to Tactile Stimulus for Potassium- Containing Dentifrices and Controls (Placebo Pastes), Measured as Threshold Force Active: Placebo: increase from rednction from Active > placebo Study Active agent' baseline (^0%) baseline (= 0%) {P value) Silverman, 1985=^ 5% KNOjd) 108% 37% PiOOOl 5% KNO3 (2) 130% PiOOOl Salvatoetal. 1992^^ 3,75% KCI (3) 114% 69% P=O.Û1 Schiff etal. 1994-° 5% KN03(4) 145% 22% P< 0.001 Silverman et al, 1994^^ 3.75% KCI (3) 101% 32% P < O.OS 3.75% KCI (7J 94% P < 0.05 Silverman et al. 1996^^ 5% KNO^CI) 99% 68% Pi O.OOi 5% KNO3 CS) 109% Pi 0 001 Schiff et al,19983' 5% KNO3 (7) 254% 110% P< 0.0001 An increase from baseline is a ir sensitivity In this Isbie. baseline is taken as 0%, thus means a doubiing of threshold. *1 = Denquel: 2 = Sensodyne F/Promise, 3 = Sensodyn ^ 4 = Sensitive/Tartar Control; 5 = Sensitivity Protection Crest: 6 = Aquafresh Sensitive; 7 ^ brand not specified.

Table 4 Percentage Reduction in Sensitivity to Cold Air Stimulus for Potassium- Containing Dentifrices and Controls (Placebo Pastes) Active: Placebo; reduction reduction Active > placebo Study Active agent* from baseline from baseline (P value) Tarbetetal, 1980^^ 5%KNO3(1) 65% 20% P=0.02 i^anochehr Pour 5% KNOjd) 63% 37% P-0.3B etal, 1984^" 5% KNO3 12) 52% P = 0.21 Silverman, 1985^^ 5% KNO, 11) 75% 40% P< 0.001 5% KNO3 121 73% P < 0 001 Salvatoetal, 1992^ 3.75% KOI (3) 66% 32% P= 0.001 Nagata et al, 1994^* 5% KNO5 17) B0% 27% P<0.01 Schiffetal, 1994^0 5% KNO3 (4) 61% 0% P< 0.001 Silverman etal, 1994^^ 3.75% KOI 13) 66% 28% P<0.05 3.75% KCI 17) 61% P< 005 Siivemian et al, 1996^^ 5%KNO3(1) 54% 30% P< 0.001 5% KNOa Í5) 48% P= 0.001 Gillametal, 1996'^ 3.75% KCH3) 51% 48% P= 0.52 West et al, 1997'^ 5% KNOa '6) 48% 53% P= 0,61 Schiffetal, 1998=' 5% KNO3 (7) 82% 40% P< 0.0001 *1 = Denqjei; 2 = Sensodyne F/Promise^ 3 = Sensodyne Fi 4 = Sensit ve/TartarCQntro;5 ^ Sensitivity Protection Crest: 6 ^ Aquafresh Sensitive; 7 - brand not specified.

Journal of Orofacial Pain 13 Orchardson/Gillam

Table S Percentage Reduction in Patients' Subjective Ratings of Sensitivity for Potassium-Containing Dentifrices and Controls (Placebo Pastes)

Active; Placebo: reduction reduction Attive > placebo Study Active agent" from baseline from baseline (P value} Tarbetelal, 1980^^ 5% KNOad) 92%^ 21%' P= 0.001 Manochehr-Pour 5% KNOjd) 54% 60% P = 0.66 etal, 19843" 5% KNOa C2) 36% P=0,81 Silverman, 1985^^ 5% KNOad) 71% 36% Pi 0,001 5% KNOa C2) 75% P< 0.001 Saivatoetal, 1992== 3.75% KCI C3) 75% 23% P = 0,001 Nagaia el al, 1994^^ 5% KNOa (7) 82% 28% P

of analysis {logit analysis: see below). The study by used a logit analysis. This method does not mea- West et al"' employed a 4-week wash-in period sure actual sensitivity levels, but assesses efficacy before the trial itself started. Whether these factors on the basis of the proportions of sensitive and could contribute to the outcome differences is not non-sensitive teeth among the study population of certain. However, the studies that did not demon- all the subjects' canine and premolar teeth at dif- strate a significant difference between tbe KNO, ferent time-points during the trial. These investiga- or KCI pastes and tbe control pastes were charac- tors fomid that the K citrate paste was significantly terized by an appreciable reduction in sensitivity more effective than a paste containing S% KNO3 by the controls, rather than any diminution of the and the control paste. All 3 pastes contained 0.8% effects of the KNO, or KCI pastes (Tables 2 to 5). Na MFP, Thus it seems likely that variations in the extent of Summary: Effects of Potassium Toothpastes. On the control/placebo responses may account for balance, the majority of trials show that potas- most of the disparities between trial outcomes. sium-containing pastes are significantly more effec- Effects of Fluoride in a Desensitizing Tooth- tive than controls in reducing DH. Formulations paste, fluoride is included in many desensitizing containing KCI and KNO3 have been accredited by pastes, but its presence is a potentially confound- the British Dental Association and the American ing variable. Topical fluorides are reported to Dental Association as effective in the treatment of reduce DH.'' Kanouse and Ash"" first reported that DH (see Tables 2 to 5). However, it is possible Na MFP in a toothpaste was an effective desensi- that at least some of the desensitizing effects of tizing agent in its own right. West et al'^ con- toothpastes may be due to the actions of confirmed the desensitizing efficacy of a paste contain- stituents other than the designated active agent, eg, ing Na MFP and reported that this "control" paste abrasive particles.'-''•'^•''-•''^ This is supported by in was no less effective in reducing dentin hypersensi- vitro data on the effects of agents on dentin perme- tivity than specific desensitizing toothpastes con- ability. Pashley et al'*'' showed that while the active taining 5% KNO3 or 8% strontium acetate. A Ingredients of several desensitizing toothpastes similar finding was reported independently by produced no significant reduction in dentin perme- Gillam et aF^ (see above). Several desensitizing ability when tested on their own, the toothpaste toothpastes containing 5.5% K citrate are avail- formulations (even when diluted by 1 part in 3) able, but there is only one published study report- produced significant reductions in dentin perme- ing the desensitizing efficacy of a toothpaste con- ability. Furtbermore, comparable reductions in taining S.3% K citrate.^^ While most other studies dentin permeability were produced by "placebo" report data for sensitivity levels, Chesters et aP^ pastes that lacked the "active" ingredient.

14 Volume 14, Number 1,2000 Orchardson/Gillam

Mode of Action has not been confirmed for human dentin. If the desensitizing effects of potassium are due to action [f potassium-containing preparations are to be potential inactivation, one might expect a transient regarded as viable desensitizing agents, what is pain when a potassium-containing toothpaste is their mode of action? It is difficult to measure first applied. To our knowledge, this phenomenon directly the effects of desensitizing agents on has not been reported for toothpastes. However, dentin perme.ibility in human teeth in situ. Thus, Creen et al,'° who used near-saturated pastes, the effects of agents on dentin permeability are noted that: "Upon application of the potassium generally carried out in vitro, nittate, most patients reported discomfort (sting- Greenhill and Pashley'" measured the effects of ing). No discomfort, or only minor discomfort, vatious desensitizing agents on the permeability was reported on application of the calcium (hydtaulic conductance) of dentin disc prepara- hydroxide.""^'''' " Although this pain could have tions. They found that 30% KNO, solution pro- arisen due to a direct action of potassium on duced no change in dentin permeability. The lntratubular nerves, it could also be due to an changes in hydraulic conductance of dentin pro- osmotic effect exerted by the very concentrated duced by other potassium salts were: 2 mol/I. K (saturated) solution.^"' carbonate: 8% reduction; 2 mol/L K carbonate + There is no direct evidence that potassium ions 2% calcium chloride: 26% reduction; 40% zinc exert their desensitizing effects directly on chloride + 20% K ferrocyanide: 38% reduction; intradental nerves when applied to human dentin. 30% K oxalate: 98% reduction. The effect of K Since it would be very difficult to test this experi- oxalate was anributed to precipitation of insoluble mentally, it is necessary to use indirect means to calcium oxalate within dentinal tubules.'' examine this hypothesis. It is well established that A Direct Action on Intradental Nerves. Since extracellular [K"^] has a profound effect on the dentin-desensitizing agents such as KNO. do not excitability of nctve and muscle cells. Increasing teduce dentin permeability by tubule occlusion,"''' extracellular [K*] decreases the resting membrane alternative modes of action have to be considered. potential (depolarization) of living cells.'^-^^ This It has been suggested that potassium ions might applies to all cells but will have the greatest conse- exert their desensitizing effects directly on intraden- quences for excitable cells (nerve and muscle), tal nerves.*''"'^ This hypothesis is based on evidence which possess voltage-sensitive ion channels. from animal experiments. Concentrated solutions Increasing [K"^J will block the action potential in of potassium salts (usually > 135 mmol/L) applied nerves because the sustained depolatization will either directly to the pulp or to deep cavities (50 to leave the voltage-gated Na* channels in a state of 100 \¡m from the pulp) in the teeth of cats or dogs relative inactivation (in effect, the membrane is produced a brief e.xcitation of inttadental nerves, held in a prolonged state of refractoriness). followed by a period of reduced excitability to Effect of Potassium Ions on Ner\'e Excitability. stimulation."' "^' Markowitz et al'" proposed that Peacock and Orchardson*" showed that the ampli- the desensitizing effects of potassium ions were due tudes of the Aß, AÔ, and C-fiber components of to increased potassium ion concentration ([K*l) in the compound action potentials of isolated nerves the extracellular fluids surrounding the intradental were attenuated when extracellular [K*"] was raised nerves. The increased [K*] causes a sustained depo- to 8 mmol/L and beyond. The conduction block larization of the nerves, resulting in inactivation of was reversible within 10 minutes of restoring the action potential generation through a mechanism [K""] to normal values (4 mmol/L). Although there such as axonal accommodation.'- There is some were no significant diffetences between the effects evidence that the desensitizing effects of K* may of KCI and KNO-, other experiments indicate that vary for different stimuli. Markowitz et al'^ the anion present can have some effect on the K- reported that while KCI decreased the responses of induced reduction in nerve excitability. It was intradental nerves to applications of hypertonic found that K citrate and K tartrate were more NaCI solutions, it was less effective in suppressing effective in blocking action potential conduction in responses to stimulation of dentin with air-blasts isolated netves than K oxalate, which in turn was and mechanical probing. One interpretation of more effective than KCI or KNOj.''* These obser- these observations is that there may be different vations are consistent with the finding that a sites or mechanisms of impulse generation for dif- toothpaste containing 5.3% K cittate was more ferent types of dentinal stimuh. effective than a 5% KNO^ paste or a control paste It is important to emphasize that this mechanism containing Na MFP in reducing numbers of hypet- is based on results of experiments on animals and sensitive teeth.^-^

Joumal of Orofscial Pain 15 Orchardson/Gillam

Diffusion of Potassium Ions Through Dentin. are applied via a toothpaste vehicie, the increase in Potassium ions applied to outer dentin will have to |K+] is likely to be transient, and the [K*] that may diffuse along the dentinal tubules to achieve ¡and be achieved is lessened by conditions tbat increase sustain) a [K*] of at least 8 mmol/L to inactivate tbe tubular fluid flow velocity or the permeability intradental nerves at the pulpal ends of dentinal of the barrier between tubules and tbe pulp. One tubules or even in the peripheral pulp. interesting finding from the model was that the It is technically very difficult to measure tbe [K""] "steady-state" [K*] at the inner end of dentinal that might be achieved within dentinal tubules in tubules could vary in cbe range of 2.6 to 19.2 vivo when concentrated solutions of potassium mmolA., depending on factors such as the perme- salts are applied to outer dentin. Therefore, alter- ability of the tubule-pulp barrier and tbe velocity native approaches have to be used. Miller et aP'' of fluid flow outward along the tubules. Tubular measured the penetration of K* through 800-|am- [K*] was lowest when outward fluid flow was thick, acid-eiched dentin discs m a 2-chamber dif- high. This raises the intriguing possibility that fusion cell. The dentin disc separated 2 chambers, tubular [K"*l could influence (or even determine) which contained phosphate-buffered saline. the excitability of intratubular nerves.^' Potassium nitrate was added to the first chamber This model perhaps oversimplifies a complex to produce an initial [K+] of 2%. The first chamber problem. Several assumptions have had to be was at atmospheric pressure, while the pressure in made, based on available information. Figures for the second chamber was raised to 20 cm H2O average fluid flow tbrough dentinal tuhules were above atmospheric to mimic the pulpal interstitial based on in vivo measurements made by fluid pressure. Samples were taken from the sec- Vongsavan and Matthews.*- However, it now ond chamber every 60 seconds for 20 minutes and seems that fluid flow rates vary considerahly potassium concentrations were measured. Miller et among tubules. Macpherson et aP' used scanning aP^ found very limited potassium diffusion across electrochemical microscopy to examine fluid flow etched but otherwise untreated discs. However, through small areas of dentin. These investigators they found greater amounts of potassium in the reported that flow was not uniform and that second chamber after the dentin discs had received much of the flow can occur through a small num- multiple treatments with the Sensitive/Tartar con- ber of tubules. If this in vitro situation also applies trol dentifrice (Colgate-Palmolive). However, the in vivo, it suggests that there would he limited molar concentrations achieved were not stated and inward K* flux along tubules with high flow rates. could not be calculated from the data presented. But there may be greater inward K* flux along Although the discs were soaked overnight in adjacent tubules with lower fluid flow rates. This deionized water after being treated with the in vitro ohservation is consistent with morpho- Sensitive/Tartar control dentifrice, the methodol- logic evidence. Dentinal tubules are not simple, ogy did not exclude the possibility that the K* col- parallel, tapered tubes of regular dimensions. The lected in the second chamber were due to release density and diameter of tubules vary at different of K* trapped within the tubules of the treated levels through dentin and at different locations in dentin rather than flowing through from the first the tooth.^•^ Tubules also exhibit varying degrees chamber. This question could be resolved by the of brancbing,*"* and tbis may help to explain the use of labeled tracers. regional variations in dentinal fluid flow rates.^' Stead et al*'^ used a mathematical model to Inward diffusion of K* along dentinal tubules is investigate the factors affecting [K^l in dentinal opposed by an outward bulk flow of tubular fluid. tubules. Although only an approximation, the Tbe diffusion rate varies with the square of the model indicated that the most important factors tubule radius, while the fluid flow varies with the affecting the steady-state [K*"] m dentinal tubules fourth power of the tubule radius. This means that were the [K'| in the mouth, the flow velocity of halving the tubule diameter will decrease diffusion tubular fluid, and the potassium permeability of by 75%, while the fluid flow is reduced by 94%. any functional diffusion barrier between the tubule The consequence is that ionic diffusion is likely to and the pulp. The model also indicated that under be greater along narrow tubules, as a result of the appropriate conditions, applications of high con- proportionately greater reduction in bulk fluid centrations of potassium ions to outer dentin can flow. This is borne out by experiment. Pashley and increase [K*] at the inner ends of dentinal tubules Matthews^-^ found that diffusion of '^^I across to > 10 mmol/L. On the basis of other experi- dentin disc preparations in the absence of any bulk ments,^^ such [K*^] would be expected to block fluid movement was not significantly affected by action potential propagation in nerves. When K* removal of a surface smear layer. However, when

16 Volume 14. Number 1. 2000 Orchardson/Gillam inward diffusion was opposed by an outward con- tal pain.*^ However, attempts to provide firm evi- vective flow generated by a hydrostatic pressure dence that odontoblasts are actively involved in sen- difference of 15 cm H,0, the reduction in diffusive sory transduction have so far been unsuccessful. flux (compared with the no-flow state) was less Guo and Davidson'*' used patch-clamping to study when the smear layer was present. isolated odontoblasts and repotted that odontoblast In principle, it seems possible to block intraden- membranes contain several types of K+ and CI" tal nerves by raising extracellular [K*], but the channels. These investigators suggest that these blocking concentrations have to be maintained channels might play an important role in dentino- after application of the desensitizing agent. genesis and perhaps in sensory transduction. They Evidence from experiments on netve excitability sutmise that stimuli might act (directly or indirecdy) indicates that potassium-induced effects are tran- on the odontoblast membrane and alter ionic con- sient and reversible.^" Unless inwardly diffusing K* dtictances. This could in turn trigger the release of can somehow be trapped at the inner ends of the various cytokines, which could evoke or modulate a dentinal tubules, any effects on intratubular nerves response from primary afferent terminals. It is of will also be transient. However, it must be empha- interest to note that cytokines (such as IFN--y, inter- sized that there is no direct evidence that K+ actu- leukin-1, tumor necrosis factor «) can stimulate the ally penetrate the dentinal tubules in the manner expression of a calcium-dependent NOS. This hypothesized when delivered into the mouth m a sequence of events is consistent with the mechanism toothpaste or other similar vehicle. hypothesized by McCormack and Davies.** Alternative Mechanisms: The Role of the Connections have been described between pulp Odontoblast. Recognizing the constraints to K* dif- nerves and cells presumed to be odontoblasts.'' fusion along the entire length of dentinal tubules, Gap lunctions have been reported between adja- McCormack and Davies^^ proposed an alternative cent odontoblasts and between odontoblasts and mechanism for K-mediated desensitization. These structures that may be nerve endings. Very investigators suggest that K+ applied to dentin recently, Ushiyama et al'- reported that stimula- could act through a second messenger system. The tion of the inferior alveolar nerve causes short- hypothesis is that K* act on the odontoblast pro- latency {1.2 to 3.5 milliseconds), all-or-none depo- cess to release nitric oxide (NO), which in turn larizations of the membrane potential of putative produces an analgesic effect by modulating noci- odontoblasts in a lower canine tooth. Tbe depolat- ceptive input through down-regulation of sensi- izations can follow stimulation frequencies of up tized nociceptors.** Nitric oxide is known to be to 300 Hz, suggesting that the linkage is not chem- ptesent in the tooth pulp, where it has diverse func- ically mediated. At present, there is no confirma- tions.*'' For example, it causes vasodilation (hence tory histologie evidence. One must beat in mind its former name: endothelium-derived relaxing fac- that such connections can operate in both direc- tot|. There is also some evidence that NO that is tions. The intradental nerves could have a trophic function a,s well as (or instead of) a sensory func- formed in response to stimulation of pulp nerves tion. It is known tbat there is a decrease in the rate may reduce the release of neuropeptides (substance of formation of secondary dentin after denervation P and calcitonin gene-related peptide [CGRP]) with capsaicin.^^ from afferent nerve terminals.*' Plasma protein extravasation from pulpal blood vessels in response to stimulation of the inferior alveolar nerve is inhibited by a substance P antagonist and a Conclusion cyclooxygenase inhibitor but is unaffected by the NO synthase (NOS) inhibitor L-NAME (N'^-nitro- Although there is evidence that toothpastes con- L-arginine methyl ester) or by a CGRP antagonist. taining K^ are more effective than minus-active There is no direct evidence for the mechanism pro- preparations in reducing dentinal hyper sensitivity, posed by McCormack and Davies,** although the K-containing preparations are not always supe- odontoblasts do contain enzymes necessary for NO rior to controls such as Na MFP, Since solutions of synthesis.**" However, it is not clear how and where potassium salts (apart from potassium oxalate) do the NO acts, and until more information is not decrease dentinal permeability, it has been sug- obtained, this hypothesis must remain speculative. gested rbat K* diffuse along the dentinal tubules to Potassium, Odontoblasts, and Sensory Trans- act on nerve terminals at tbe pulpal ends of the duction? Perhaps because of their relationship to tubules. Predictions from computer models indi- dentinal tubules and the intradental nerve termi- cate that K*^ can diffuse along tubules in sufficient nals, the odontoblasts have been implicated in den- amounts to block netve conduction. However, any

Joumal of Orofacial Pain 17 Orchardson/Gülsm

effects will be transient, unless tbe higb intratubu- 17. Greenhill JD, Pashley DH. The effects of desensitizing lar [K*^| can be maintained. At present, tbcre is no agents on rhe hydraulic conductance of human dentin in convmctng evidence that desensitizing preparations vitro, J DentReb 1981;60;686-698. based on KCI, KNO,, and K citrate act in tbe 18. Mordan NJ, Barber PM, Gillam DG. The dentine disc, A review of its applicability as a model for the in vitro test- manner proposed. It is possible tbat any desensitizing of dentine. J Oral Rehabil 1997;24;J48-156, ing actions of toothpastes may be due to tubule- 19. Smith BA, Hanssun RE, Caffesse RG, Bye FL, Evaluation occluding effects of constituents other than the of dipotassium oxalate in the treatment of root hypersen- potassium salts. sitivity [abstract 1733J. J Dent Res l9Sli;67;329. 20. Holborow DW. A clinical trial of a potassium oxalate system in the treatment of sensitive root surfaces. Arch Oral Biol 1994;39(suppl);134S. References 21. Muzzin KB, Johnson R. Effects of potassium oxalate on denrin hypersensitivity in vivo. J Periodonrol 1989; 1. Addy M. Clinic^Ll aspects of Jeiitinc hyperseiisiriviry. Prot 6Û;151-1S8. Finn Dent Soc 1991;SS(suppl 1);4O7-4I2. 22. Gillam DG, Coventry JF, Manning RH, Newman HN, 2. Närhi MVO. The characteristics of intradenral sensory Bulman JS. Comparison of two desensitising agents for the units and their respotises to scimulatiun. J Dene Res treatment of cervical dentine sensitivity. Endod Dent Tranmatol 1997;l3;36-39. 3. Matthews B, Vongsavan N, Interactions between neural 23. Coûley RL, Sandoval VA. Effectiveness of potassium and liydrodynamic mechanisms m dentine atid pulp. Arch oxalare treatment on denrin hypersensitivity. Gen Dent Oral Biol 1994;39(sijppl);87S-95S. 4. Närhi M. Yamamoto H, Ngassapa D, Hirvonen T. The 24. Giienin ME, Scheldt MJ, O'Neal RB, Strong SL, Pashley neuropliysiological basis and rhe role of inflammatory DH, Horner JA, Van Dyke TE. An in vivo study on dentin reacciotis in dentine hypersensitivity. Arth Oral Biol sensitivity; The relation of dentin sensitivity and patent 1994;39(suppl):23S-30S. dentin tubules. J Periodonrol 1991;62;668-673. 5. Matthews B, Andrew D, Amess TR, lkeda H, Vongsavan 25. Holland GR, Närhi MN, Addy M, Gangarosa LP, N. The functional properties of intradeiital nerves. In; Orchardson R. Guidelines for the design and conduct of Shimono M, Maeda T, Suda H, Takahashi K (eds). Dentin/ clinical trials on denrin hypetsensitivity. J Clin Periodontol Pulp Comple;*. Tokyo; Quintessence, 1996:146-153. 1997;24;808-813. 6. Orchardson R, Gangarosa LP Sr. Hypersensitive dentine; 26. Tarbet WJ, Silverman G, Stolman JM, Frararcangelo PA. Biological hasis of therapy. Arch Oral Biol 1994;39 Clinical evaluation of a new trearment for dentinal hyper- |stippl);lS-153S. sensitivity. J Periodontol 1980;51;535-Í40. 7. Ling TYY, GLllam DG, The effectiveness of desensitising 27. Collins JE, Gingold J, Stanley II, Simring M. Reducing agents for the trearment of cervical dentine sensitivity dentitial hypersensirlvity with strontium chloride and (CDS)—A review. Periodont Abstr 199é;44;5-12. potassium nitrate. Gen Dent 1984;32:40-43. 8. Calvo P. Treatment of sensitive dentine. Dent Cosmos 28. Silverman G. The sensitivity-reducing effect of brushing 1884;26;139-141, wirh a potassium nitrate-sodium monofluorophosphare 9. Hodosh M. A superior desensitizer—potassmm nitrate, J dentifrice. Gompend Contin Educ Dent 1985;6;131-133, Am Dent Assoc )974;S8;831-832. 136. 10. Green B|, Green ML, McFall WT. Calcium hydroxide and 29. Nagata T, Ishida H, Shinohara H, Kasahara S, Wakano Y, potassium nitrate as desensitising agenis for hypersensitive Daigen S, Troullos FS. Ghnical evaluation of a potassium root surfaces. J Periodontol 1977;48:667-672. nitrate dentifrice for the treatment of dentinal hypersensi- 11. Reinhart TC, Killoy WJ, Love J, Overman PR, Sakumura tivity, J Glin Periodontol 1994;21;217-221. JS. The effectiveness of a patient-applied desensitising gel, 30. Schiff T, Dotson M, Cohen S, de Vizio W, McGool J, A pilot study. J Glin Periodontoi 1990;17;123-127. Voipe A. Efficacy of a dentifrice containing S% potassium 12. Zappa U. Self-applied treatments in the management of nitrate, 1.3% soluble pyrophosphate, 1,5% PVM/MA co- dentine hypersensitivity. Arch Oral Biol I994;39(suppl|; polymer and 0.243% sodiuni fluoride in a silica base on 107S-112S. dentinal hypersensitivity. J Clin Dent 1994;5|special 13. Gillam DG, Bulman JS, Jacksoii RJ, Newman HN. issue) ;87-92, Efficacy of a potassium nitrate motithwash in alleviating 31. Schiff T, Dos Santos M, Laffi S, Yoshioka M, Baines £, cervical dentine sensitivity (GDS). J Clin Periodonrol McCool J, de Vizio W. Efficacy of a dentifrice containing 199éa;23;993-997, 5% potassium nitrate and 1500 ppm sodium monofluo- 14. Yares R, West N, Addy M, Marlow I, The effects of a rophosphate in a precipitated calcium carbonate base on potassium citrate, cetylpyndinium chloride, sodium fluo- dentinal hypersensitivity, J Chn Dent 1998;9;22-25. ride mouthrinse on dentine hypersensitiviry, plaque and 32. Silverman G, Berman E, Ilanna CB, Salvato A, gingivitis. A placebo-controlled study. J Glin Periodontol Fratarcangelo P, Bartizek RD, et al. Assessing the efficacy 1998;25;813-820. of three dentifrices in the treatment of dentinal hypersensi- 15. Gillam DG, Bulman JS, Jackson RJ, Newman HN. tivit)'. J Am Dent Assoc 1996;127;191-201. Comparison of 2 desensitising dentifrices wirh a commer- 33. Tarbet WJ, Silverman G, Eratarcangelo PA, Kanapka JA. cially available fluoride dentifrice in alleviating cervical Home treatment for dentinal hypersensitivity; A compara- dentine sensitivity. J Periodonrol 1996b;67:737-742, rive srudy, J Am Dent Assoc 1982;105;227-230. 16. West NX, Addy M, Jackson RJ, Ridge DB. Dentine hyper- 34. Manochehr-Pout M, Bhat M, Bissada N. Clinical evalua- sensirivity and the placebo response, A comparison of the rion of two potassium nitrate toothpastes for the treat- effect of strontium acetate, potassium nitrate and fluoride ment of dental hypersensitivity, Periodont Case Rep toothpastes. J Glin Periodonto! 1997;24;209-215. 1984;6;25-30.

18 Volume 14, Number 1,2000 Orchardson/Gillam

35. Cheîters R, Kaufman HW, Wolff MS, Huntington E, 54. Anderson DJ, Matthews B. Osmotic stimulation of human Kleinherg l. Use of multiple sensitivity measurements and dentine and the distribution of dental pain thresholds. iogit statistical analysis to assess the effectiveness of 3 Arch Oral Biol 1967; 12:417-426. pota SSI um-citrate-contain ing dentifrice in reducing denti- 55. Adrian RH, The effect of internal and external potassium nal hypersensitivity. J Clin Periodontol I992;iy:256-261, concentration on rhe membrane potential of frog muscle. J 36. Person P, Demand EE, Koltun L, Spindel LM. A micropro- Pliysiol 1956il.î.î;631-65S, cessor temperature-controlled air delivery system tor denri- 56. Hoilgkin AL, Horowitz P. The influence uf potassium and nai hypersensitiv i t>'testing. Clin Piev Dent 1989;! l;3-9. chloride ions on the membrane potential of single muscle 37. Ayad F, Berta R. De Vizio W, McCool J, Pétrone ME, fibres. J Pbysiol 1959^148:127-160. Volpe AR. Comparative efficacy of two dentifrices con- 57. Peacock JM, Orchardson R. Effects of potassium ions on taining 5% potassium nitrate on dentinal iensitivity: A action potential conduction in A- and C-fibers of rat twelve week clinical sruJy. J Clin Dent l-i'MuilspccijI spinal nerves. J Dent Res 1995;74:634-641. issuelT97-101. 58. Peacock JM, Orchardson R, Action potential conduction 38. Salvato AR, Clark GE, Gingold J. Currci FA. Clinical block of nerves in vitro by potassium citrate, porassium effectiveness nf a dentifrice containing potassium chloride rartrate and potassium oxaiate. J Clin Periodonrul as a desensitizing agent. Am J Dent I992;5Í303-306. l!l99;26:33-37. 39. Silverman G, Gingold J, Curro FA. Desensitizing effect of 59. Miller S, Gaffar A, Sullivan R, Hui R, Truong T, Stranick a potassium chloride dentifiice. Am J Dent ]994;7:9-l2. M. Evaluation of a new dentifrice for the treatment of sen- 40. Gillam DG. Clinical Trial designs for testing of products sitive teeth. J Clin Dent 1994;5¡special is5üe¡:71-79. for dentine hypersensitivity—A review. Periodont Absrr 60. Stead WJ, Orchardson R, Warren PB. A mathematical 1997;45:37-46. model of potassium ion diffusion in dentinal tubules. Arch 41. Kanouie MC, Ash MM. The effectiveness uf a sodium Oral Biol 1996;41:679-687. m 0 no fluoro-phosphate dentifrice on dem a I hypersensitiv- 61. Orcbardson R. Strategies for the management of hyper- ¡t}'. J Periodontol 1969;40:38^0. sensitive reeth. In: Shimono M, Maeda T, Suda H, 42. Addy M, Mostafa P. Dentine hypersensitivity. [I, Effects Takahashi K (eds). Denrin/Pulp Complex. Tokyo: produced by the uptake in vitro of toothpastes onto den- Quintessence, 1996:85-89. tine. J Oral Rehabil J9S9;lé:ji^S. 62. Vongsavan N, Matthews B. Fluid flow through cat den- 43. Ling TYY, Gillam DG, Barber PM, Mordan NY, Critchell tine in vivo. Arch Oral Biol 1992;37:I75-185. J. An invesiigalion of potencial desensitising agents in the 63. Macpherson JV, Beeston MA, Unwin PR, Hughes NP, denrine disc modei. A scanning electron microscopy study. Lirtlewood D. Scanning electrochemical microscopy as a J Oral Rehabil 1997;24d91-203. probe of local fluid flow rhrougli porous solids. J Chem 44. Pashley DH, O'Meara JA, Kepler EE, Galloway SE, Soc Faraday Trans 1995i91i 1407-1410. Thompson SM, Stewart FP. Dentine permeabilitj'. Effects 64. Miör lA, Nordahl I. The density and branching of dentinal of desensitising dentifrices in vitro. J Petiodonrol tubules in human teeih. Arch Oral Bioi 1996;41:401^ 12. 1984;55:S22-525. 65. Pashley DH, Matthews WG. The effects of outward 45. Pashley DH. Dentine permeability, dencine sensitivity and forced convective flow on inward diffusion in human den- treatment through tubule occlusion. J Endod 1986; tine in vitro. Arch Oral Biol 1993;3 8:577-582. 12:465^74. 66. McCormack K, Davies R, The enigma of potassium ion in 46. Markowitz K, Kim S, Hypersensitive teeth. Experimental the management of denrine hypersensitivity: [s nitric oxide studies of dentinal desensitising agents. Denr Clin North rhe elusive second messenger? Pain 1996;68:5-I1. Am 1990;34:491-50L 67. Olgarr LF. Neurogenic components of pulp inflammation. 47. Orcbardson R. Ion sensitivity of intradental nerves in the In: Shimono M, Maeda T, Suda H, Takahashi K (eds). cat. In: Anderson DJ, Matthews B (eds). Pain in tbe Dentin/Pulp Complex. Tokyo: Quintessence, 1996: Trigeminal Region, Amsterdam: Elsevier/Norrb-Holland, 169-175. 1977:77-82. 68. Kerezuudis NP, Olgarr L, Fried K. Localisation of 48. Närhi M, Haegarstam G. lntradental nerve activity NADPH-diaphorase activity in the dental pulp, periodon- induced by reduced pressure applied to exposed dentine in tium and alveolar hone of the rat. Histocheniistry the car. Acra Physiol Scand 19S3;119:3S1-3B6. 1993;100:319-322. 49. Kim S, Hypersensitive teeth: Desensmsation of pulpal sen- 69. Anderson DJ, Hannam AC, Matthews B. Sensory mecha- sory nerves. J Endod 198é;12-482^85. nisms ill mammalian teerh and tlieir supporting structures. 50. Markowirz K, Bilorto G, Kim S. Decreasing intradenral Physio! Rev 1972;5O:17]-I95, nerve acriviry in the cat with potassium and divalent 70. Cuo L, Davidson RM. Porassium and chloride channels in carions. Arch Oral Biol 1991;36:l-7. freshly isolated odontoblasts. J Dent Res 1998;77:341-350. 51. Orchardson R, Peacock JM. Factors affecting nerve 71. Matthews B, Holland GR. Coupling between nerves in excitability and conduction as a basis for desensitising teeth. Brain Res 1976;98:354-358. dentine. Arch Oral Biol 1994;39(suppl):81S-85S. 72. Ushiyama J, Wanachantararak S, Li F, Matthews B. 52. Aidley DJ, The Physiology of Excitable Cells, ed 3. Functional coupling berween odontoblasts and nerves in Cambridge: Cambridge University Press, 1989:50-51. the cat labbtract 947].' J Dent Res 1998;77;750, 53. Markowir^ K, Nârhi M, Ngassapa D, Kim S. Tbe effect of 73.Jacobsen EB, Heyeraas KJ. Effect of capsaicin treatment KCI on intradental nerve activity evoked by hypertonic or inferior alveolar nerve resection on dentine formation NaCI solution and natural stimuli ¡abstract 1492J. J Dent and calcitonin-gcne related peptide- and substance P- Res 1990;69(special is$uel:295. immunoreactive nerve fibres in rat molar pulp. Arch Oral Bioll99é;41:112I-1131.

Journal of Orofacial Pain 19