Several investigators have demonstrated that intradental A-type (AB and AD) nerve fibers are responsible for the sensitivity of dentin and that the endings of the responding fibers are located in the pulp-dentin area of the tooth. The exact mode of transmission of stimuli (e.g., thermal, chemical, mechanical, etc.) across dentin, however, is still unclear, although several hypotheses have been proposed. These include direct nerve stimulation, dentinal receptor (transducer/modulation), hydrodynamic, and direct ionic diffusion hypotheses. Currently, the most accepted mechanism of intradental nerve activation associated with dentin sensitivity appears to be hydrodynamic in nature, although alternative mechanisms of transmission (e.g., direct ionic diffusion) cannot be ruled out. Recent investigations (in the cat), however, appear to provide evidence substantiating the hydrodynamic hypothesis.
Key words: innervation, dentin, odontoblast, mechanisms of action, hydrodynamic theory, direct ionic diffusion
According to Narhi and co-workers, intradental A-type nerve fibers (see classification of nerve fibers) are responsible for the sensitivity of dentin and the endings of the responding fibers are located in the pulp-dentin area. The exact mode of dentin sensitivity, however, is still unclear, although several hypotheses have been proposed. Currently, the most accepted mechanism of intradental nerve activation associated with dentin sensitivity appears to be hydrodynamic in nature.
The basic neuroanatomy of the dental pulp and dentin has been reviewed recently. Nerve fibers Pulp and Dentin entering the teeth have been identified histologically as myelinated A-fibers and unmyelinated C-fibers. These fibers are grouped in bundles and enter through the apical foramina of the teeth, passing through the radicular to the coronal pulp where they fan out and diverge into smaller bundles. Nerve divergence continues; individuals-fibers within small bundles lose their myelin sheath and divide repeatedly before finally ramifying into a plexus of single axons known as the subodontoblastic plexus or plexus of Raschkow. The exact function of this plexus is unknown, as is the changing configuration of the plexus with dentin formation." From this plexus nerve fibers are distributed toward the pulp-dentin border with terminals showing a characteristic bead-like structure. Gunji has studied the distribution of nerve terminals arising from the sub-odontoblastic plexus in human molar teeth and classified four types of nerve endings, according to where they terminated. This has been summarized by Trowbridge.
These simple pulp fibers extend from the sub-odontoblastic nerve plexus to the odontoblast layer but do not reach the predentin.
These fibers extend to the odontoblast/predentin border or enter the predentin. Gunji observed that some of these fibers ran straight or spiraled through a dentinal tubule along with an odontoblast process; others ran diagonally along the odontoblast/predentin border or within the predentin. Other fibers looped back toward the odontoblast layer.
These fibers reach the predentin and undergo terminal ramification with multiple branches and multiple ending-like enlargements on each branch. The area covered by a single such terminal complex has been estimated to exceed 100,000 um2 in some instances. Penetration of this terminal type into dentin is limited to several um.
These fibers pass through the predentin without branching and enter the dentin through the dentinal tubule. The penetration is limited to approximately 100 um.
Of the four types, the marginal fibers were the most numerous and the dentinal fibers the fewest.
Observations from early histological studies appeared to indicate that dentin was directly innervated. These observations, however, were open to conjecture. The use of silver impregnation techniques affected other structures such as reticular and collagen fibers as well as dentinal nerve fibers. Bernick overcame this particular difficulty by the use of enzymes to digest collagen. Difficulties in interpretation also resulted from the limited resolution of the light microscope and from other technical difficulties in the preparation of histological specimens. Later studies showed the presence of nerve-like fibers in dentin. Fearnhead observed fine-beaded fibers extending for a short distance into some but not all tubules. Penetration of fibers into dentin was limited to a few micrometers for most fibers, although some appeared to penetrate as far as 150-200 um. Several investigators have suggested the presence of nerve structures in dentinal tubules. Frank and Arwill, using electron microscopy, described such structures as terminal axons and sensory receptors. These observations were confirmed by nerve section studies. More recently, sensory nerves were identified in dentinal tubules by autoradiography techniques.
There is considerable variation in the number of dentinal nerve fibers from individual to individual and from tooth to tooth. Both Avery and Lilja demonstrated that only approximately 25-27% of the dentinal tubules (cuspal dentin) has associated nerve fibers. The coronal two-thirds of dentin contained no neural structures in the dentinal tubules. According to Lilja, all the available evidence suggested that all neural structures observed were confined to the predentin and the most pulpal dentin, and in the main, dentin was largely devoid of nerve fibers. Lilja also found regional differences in the extent of dentin innervation.
Ten-Cate and co-workers, in commenting on the role of the odontoblast and the extent of the odontoblast process within the dentinal tubule, suggested that the conflicting evidence provided by the differing technologies of scanning and transmission electron microscopy may have accounted for differences in interpretation. They concluded that the odontoblast cell process extended only to the amelodentinal junction.
Holland, however, stated that this position is still unclear. According to Holland, recent improvements in fixation and immunohistochemical techniques may have provided structural evidence of odontoblastic processes in peripheral dentin. Szabo and co-workers have shown that a smooth lining the lamina limitans, runs the whole length of the tubule. Other investigators, using polyclonal and monoclonal antibody labeling techniques, have demonstrated components of the cytoskeleton in peripheral dentin. Several investigators have suggested that under previous methods of fixation the odontoblast process may have retracted (from its full length in the tubule) into the inner third of dentin. Holland also has observed (in the cat) that odontoblast processes vary in length and it is possible that at some sites the process may be as long as the dentinal tubule. This supposition, if correct, could explain why some areas of exposed dentin are sensitive, while others are not, i.e., sensitive areas have tubules where the odontoblast process is closer to the exposed surface, whereas in nonsensitive areas the process is confined to the inner one-third of the dentin. Takahashi, however, stated that evidence for an artifact of shrinking due to fixation was far from certain and doubted whether these new approaches to resolve the problem had made any progress.
Nerve fibers have been classified according to their conduction velocity and axon diameter into A- (Aa, AB, Ay, and Ad), B- and C- fiber types.
Aa fibers have a diameter of 12-20 um, and a conduction velocity of 70-120 m/sec. The primary function of these fibers is one of proprioception. AB fibers have a diameter of 5-12 um with a conduction velocity of 30-70 m/see and are responsible for the transmission of touch and pressure. Ay fibers have a diameter of 3-6 um with a conduction velocity of 15-30 m/s and are responsible for motor function to the spinal nerves. Ad fibers have a diameter of 2-5 um with a conduction velocity of 12-30 m/see and are responsible for the transmission of pain, temperature, and touch. Fibers in the second group, B-fibers, have a diameter of 1-3 um with a conduction velocity of 3-15 m/see and are responsible for preganglionic autonomic function. C-fibers are unmyelinated and have a diameter of 0.2-2 um with a conduction velocity of 0.5-2 m/sec. Their functions include postganglionic sympathetic pain and possibly heat, cold, and pressure.
Dental pulp is innervated by both myelinated and unmyelinated fibers. By tooth eruption, both myelinated and unmyelinated nerves have reached the odontogenic regions and lie close to the odontoblast. Recent electrophysiological investigations on intradental nerves of experimental animals confirm histological evidence that two fiber groups, A- and C-, exist and provide both fast and slow conduction. The groups are functionally different.
According to Narhi and co-workers, it would appear that A- fibers are responsible for the sensitivity of dentin (dentinal pain). Most are Ad fibers, whereas C-fibers respond when external irritants (e.g., chemical agents) reach the pulp (pulpitis).These investigators also have demonstrated that there are other intradental nerve units that have conduction velocities above the range of the Ad fibers. These have been classified as AD fibers and appear to respond in the same way as Ad fibers to drilling probing of dentin, and air blast, which would indicate that both Ad and AB fiber units belong to the same functional group. AB fibers may mediate non-painful sensations induced by low-intensity electrical stimulation of human teeth.
Pashley and Parsons suggested that the mechanism of dentinal sensitivity transmission can be of Stimulus classified according to three main hypotheses: (1) nerve endings or nociceptors that respond Transmission directly when the dentin is stimulated, located throughout the dentin; (2) odontoblasts, being chemically or electrically related to nerves, function when depolarized as receptors generating nerve impulses; and (3) stimuli applied to dentin producing a displacement of dentinal tubule contents which could excite mechanosensitive nerve endings near the pulpal end of the tubules (hydrodynamic mechanism).
Several investigators have previously maintained that other mechanisms of pulpal sensory nerve activation (i.e., direct ionic diffusion) may be responsible for the transmission of stimuli across dentin. Other investigators have also reported that chemical agents (e.g., 3M NaCl) did not elicit intradental nerve activity when applied in shallow dentinal cavities, but did cause excitation of Ad fibers when applied in deep cavities, the most probable mechanism for this mode of action being direct ionic diffusion. From these studies Markowitz et al. concluded that chemical agents were able to diffuse through the dentinal tubules and directly alter the extracellular fluid environment of the intradental nerves (1) by changing the extracellular fluid environment and altering the critical level for firing action potentials, (2) by directly altering membrane properties through a specific chemical interaction leading to a change in permeability, and (3) by changing the microcirculation of the pulp.
Anderson and co-workers postulated that if dentin was directly innervated, then chemical stimuli to the exposed sensitive dentin surface should cause pain. Application of algogenic (pain-inducing) substances such as potassium chloride, acetylcholine, 5-hydroxytryptamine, and histamine failed to elicit a response; whereas when applied directly to exposed pulpal tissue, an immediate response was elicited. Similarly, topical anaesthetic solution when applied to the exposed sensitive dentin did not decrease sensitivity. Anderson and Naylor proposed two possible explanations: 1. There were no nerve elements in dentin. When pain was evoked it was due to stimulation of receptor mechanisms in the pulp by a disturbance transmitted through the tubules by non-neural means. 2. There are receptor mechanisms in dentin that could be stimulated indirectly, but cannot be reached by direct stimulation from chemical agents because of some barrier to diffusion in the tubules.
Naylor, however, observed that the very fast pain (cold) reaction times following thermal stimulation did, in fact, suggest the presence of a receptor located in dentin. Naylor later demonstrated that disruption of the odontoblast layer under a cavity did not block pain sensation following cold stimulation.
Application of sugar and calcium chloride solutions with high osmotic pressures did, however, produce pain in dentin, although this does not prove that a receptor mechanism is present in dentin, since nerves in the pulp may have been stimulated.
Recent autoradiography studies of intradental nerves have demonstrated that nerve fibers in the pulp/dentin border area are injured by dentinal stimulation, with a 50% reduction in the number of innervated dentinal tubules and, in some instances, loss of the nerve fibers in dentin. These results suggest that the existence of nerve fibers in dentin is not a necessary prerequisite for its sensitivity, which also supports the evidence of Lilja that root dentin contains no intratubular nerves, but nevertheless is very sensitive. Several investigators, however, have reported that following injury to dentin (in the rat molar), nerve fibers rapidly sprout under the injured cervical dentin provided the odontoblast layer is not destroyed. These sprouting calcitonin gene-related peptide immunoreactive fibers (CGRP-IR) can temporarily innervate dentin of the root. This effect, however, appears to diminish within 21 days. They suggested that exposed hypersensitive dentin may have more nerve fibers in the underlying pulp and dentin.
Kramer concluded that a lack of correlation between disturbance of tubule contents and pain experience indicated that dentin sensitivity cannot be explained in terms of movement of tubule contents.
Proponents of the dentinal receptor mechanism hypothesis have suggested that the odontoblast has a special sensory function (although this receptor does not have to be the odontoblast), and that a functional complex with the terminal sensory nerve endings in close proximity to the odontoblast layer acts as an excitatory synapse. These so-called specialized junctional complexes were concluded to be a unique type of 'neurosensitive complex.' Arwill demonstrated cell projections in the predentin, which he called associated cells, but not in dentin itself. The presence of tight junctions has also been described. Several investigators, however, have failed to establish the presence of any synaptic junction or special form of connection between odontoblast process and nerve endings, although an intimate contact between axon and odontoblast has been noted. Morphological evidence of a synaptic relationship between odontoblast and sensory nerve endings is lacking.
Gunji hypothesized that free sensory nerve endings may in some way couple with the odontoblast process to form a mechanoreceptor complex capable of being stimulated when the odontoblast is mechanically deformed. This fails to explain why dentin continues to be sensitive following experimental destruction of the odontoblast layer. Lundy and Stanley showed that after odontoblast degeneration, clinical sensitivity persisted. Dentinal sensitivity persists following the degeneration of both odontoblasts and intratubular nerve fibers in the inner third of dentin. Such studies appear to contradict the hypothesis that odontoblasts act as a dentinal receptor mechanism. Several investigators have stated that it was unlikely that the odontoblast could perform the function of a special sensory receptor cell, which at the same time functioning as the specialized formation cell of dentin.
Various histochemical and electrophysiological studies have investigated the possibility of a synaptic connection between terminal sensory nerve endings and odontoblast processes (transducer theory). In order to substantiate this theory, the presence of a neurotransmitter substance, such as acetylcholine, would have to be demonstrated by evidence of acetylcholinesterase activity in the dentin. Avery and co-workers have shown that odontoblast protoplasmic extensions were cholinesterase positive. Ten-Cate and Shelton demonstrated cholinesterase activity in both myelinated and non-myelinated nerve fibers of the pulp, but not close to, or in, odontoblasts or their processes. They concluded that if the transmission of impulses associated with dentinal sensitivity was via a dentinal receptor mechanism, then there was no evidence to suggest that these impulses were mediated by cholinergic activity.
Several investigators have suggested that nerve impulses in the pulp may be modulated by polypeptides, such as plasma kinins and Substance P (modulation theory). Most studies, however, have failed to demonstrate any morphological evidence of a synaptic relationship between odontoblasts and sensory nerve endings. Any direct effect of an external stimulus on pulpal nerves would also be unlikely due to the insulating properties of dentin which, apart from the predentin and the most pulpal aspect, is largely devoid of nerve fibers. These findings would, therefore, lend support to an indirect stimulatory mechanism.
Several investigators, however, have claimed that recorded electrical activity from dental nerve fibers indicated the presence of receptors in dentin. Arwill and co-workers also reported that when electrophysiological recordings were made on teeth which had had the inferior alveolar nerve resected, no impulse activity was recorded; whereas teeth on the control side with an intact nerve responded to locally applied stimuli. These investigators also observed associated cell degeneration on the nerve resected side. They postulated that the associate cell described in human teeth was actually a sensory neuron. This observation would appear to confirm the earlier electrophysiological studies of Scott and co-workers that gave some credence to the concept of a dentinal receptor mechanism. According to Anderson and co-workers cited by Matthews, however, the evidence from the studies of Scott and co-workers depended partly on their interpretation of the shapes of the recorded wavelengths as well as on the fact that these investigators were unable to record any activity until dentin was removed to within 100-200 um of the pulp where nerve fibers are known to be present. The possibility that the impulse activities recorded were those of pulp nerves, therefore, cannot be excluded. Matthews also reported that the response to stimulation recorded directly from an intact pulp was similar to that from the overlying dentin. Both Winter and co-workers and Matthews failed to demonstrate any recorded impulse activity which could be attributed to the odontoblast. Similarly, other studies noting the low-membrane potential of the odontoblast also failed to demonstrate any recorded impulse activity. Horiuchi and Matthews also demonstrated that the recording system of Scott and coworkers could cause artifacts in the recorded activity. Furthermore, Horiuchi and Matthews observed that a recording electrode (Ag/AgCl) in contact with dentin may be capable of recording activity from beyond the immediate subadjacent tubules. Evidence from these studies, therefore, would indicate that the odontoblast does not possess the properties of a sensory receptor.
The exact mechanism of impulse transmission remains controversial. Mjor and Pindborg have stated that pulp and dentin sensation is characterized by being limited to pain only, irrespective of the initiating factor. According to Berman, however, there is no direct support for any specialized terminal nerve receptors for hot, cold, electrical, osmotic, dehydration, or chemical stimuli in dentin, although several investigators have demonstrated that, once the impulses reached the pulp, there were definite heat- and cold-sensitive nerves present. Several investigators have shown that individual dental nerves (neurons) in animals respond to several different types of stimuli such as drilling probing air drying and hyperosmotic solutions. A study in humans also reported that cold perception was evident following application of a cold stimulus, although when these investigators anesthetized the gingivae, this relationship was reduced. Jyvasjarvi and Kniffki also reported that no sensation other than pain was perceived.
Dentin is composed of hollow tubes containing a fluid or semifluid material. Neither Neil nor Kramer, however, was convinced that dentinal fluid movement was an acceptable explanation for the generation of pain. Gysi proposed that movement in dentinal canaliculi in either direction resulted in a sensation of pain. Fish had also proposed the idea of a fluid within the dentinal tubules, apparently extracellular. Ishikawa postulated that the pulpal lymph flow was continuous with that of the dentinal tubule fluid, even though he failed to observe any pulpal Iymphatics (in the dog. Stanley observed that free fluid made up about 2% of enamel volume and 25% of that of dentin. Kramer considered the dentinal tubule wall to be a relatively rigid structure. Johansen and Parks also observed that the walls were considerably more mineralized than the rest of the dentin. The diameter of these tubules was 2.5 um at the pulpal end and 0.9 um peripherally. Brannstrom reasoned that the conical shape of the dentinal tubules, together with the movement of fluid by capillary attraction, should obey the same physical laws as liquids in glass capillary tubes (Poiseuille¹s Law).The movement of fluid within the tubule was calculated to be about 2-4 micrometers per second. Stanley also demonstrated that mobility of the fluid was high. Low hydrostatic pressures of 2 kg/cm3 were also observed to elicit pain and to cause incremental flow of dentinal fluid toward the pulp as opposed to the slow outward flow that normally appears to occur. This spontaneous rate of outward fluid movement, which flows down a hydrostatic pressure gradient from the pulp, is apparently too slow to activate mechnoreceptors. Johnson and co-workers also observed that in fractured dentin with exposed tubules, tubule contents could be emptied about 10 times a day.
Brannstrom suggested that the displacement of tubule contents, if rapid enough, could deform nerve fibers in pulp or predentin, or damage odontoblast cells; both effects appear capable of producing pain. More recently, this definition has been refined to state that minute fluid shifts, either dentinal fluid or tubule contents, across dentin in either direction, in response to tactile, thermal, or osmotic (chemical) stimuli, can stimulate mechanoreceptors in or near the pulp, which, in turn, excite sensory nerves to cause pain
According to Pashley, the hydrodynamic theory of dentin sensitivity as proposed by Brannstrom is based on the premise that sensitive dentin is permeable throughout the length of the tubules.
Currently, most investigators accept that dentin sensitivity is due to hydrodynamic fluid shifts which occur across exposed dentin with open tubules. This rapid fluid movement, in turn, activates the mechanoreceptor nerves of the AB and Ad classes in the pulp.
In a series of experiments, Brannstrom and other investigators demonstrated that fluid shifts occurred through the dentinal tubules when pressure and dehydration, as well as thermal stimuli were applied to dentin.
The effect of pressure: The effect of pressure in teeth with cavity preparation made into dentin has been evaluated by Brannstrom. Following a decrease in pressure, an immediate pain response was elicited which persisted for as long as there was decreased pressure. Histological examination showed odontoblast nuclei in the tubules. Brannstrom concluded that these effects were probably due to intense evaporation from the dentinal surface. The dislocation is probably due to aspiration of the odontoblast into the dentinal tubules in connection with the capillary fluid flow. Dislocated odontoblast nuclei also have been demonstrated in the dentinal tubules under stimulated dentin. Kramer observed this, but failed to correlate the incidence of disturbance and pain experienced.
Brannstrom and Astrom postulated that rapid fluid shifts might activate nerves located at some distance from the tubules corresponding to the exposed dentin.
The effect of dehydration: Several investigators have demonstrated that the placement of dry absorbent paper on exposed dentin elicited a painful response, whereas with wet paper, no pain was experienced. Scratching of the exposed dentin with a sharp probe or by dry chiseling also elicited a painful response. These procedures could cause the removal of dentinal fluid from the exposed dentin surface and by capillary action elicit an outward flow of tubule contents from the pulp, stimulating the odontoblast structure and causing pain. Recently Vongsavan and Matthews have demonstrated that gentle probing caused inward movement of fluid.
Hypertonic solutions such as sugar and calcium chloride also elicit pain by the same effect of dehydration of the dentinal surface. Bender has shown that the discomfort subsides when the irritant is diluted. This effect, too, can be explained by dentin tubule fluid movements, since fluids of a relatively low osmolarity (e.g., dentinal tubule fluid) will tend to flow toward solutions of higher osmolarity. When isosmotic solutions are applied, no stimulus is perceived. According to Haegerstam and co-workers, the receptors of the tooth (in the cat) are not chemoreceptors, but probably mechanoreceptors. This hypothesis is supported by the recording of nerve impulses following application to dentin of stimuli known to create fluid movements in tubules.
Several investigators also have shown that intradental A- nerve fiber units in animals respond to several different types of stimulus affecting dentin such as drilling probing air drying and hyprosmotic solutions. These stimuli induced pain when applied to human dentin. Lilja demonstrated sensory differences between crown and root dentin using dry absorbent paper, air blast, and calcium chloride solution. Pain in crown dentin was sharp and shooting that in root dentin, dull and often of longer duration. Acid etch treatment also has been utilized in studies of dentin sensitivity. This treatment is known to remove any drilling debris contributing to the smear layer and to open the dentinal tubules. Acid etching also may enable the stimulus to evoke intratubular fluid movements, which in turn makes the intradental nerve units more responsive. This factor may not have been accounted for in studies that failed to demonstrate a response to osmotic stimuli. It should also be noted that the early works of Anderson and Brannstrom were essentially physiological studies of coronal dentin, rather than of cervical dentin sensitivity per se. Anderson and Matthews also cautioned against extrapolating their observations on non-carious healthy coronal dentin to sensitive root dentin. Another factor is the absence of a smear layer on the exposed root surface of teeth in patients complaining of sensitivity.
The effect of thermal changes: According to Berman, the perception of acute thermal stimulation can be explained by the hydrodynamic theory. When a cold stimulus was applied to dentin, it was observed to cause a contraction of tubule contents, which in turn resulted in a rapid outward movement of fluid away from the pulp. Conversely, when heat was applied to dentin, expansion of tubule contents occurred with a subsequent increase in pressure, which resulted in a rapid inward movement of fluid toward the pulp. In a series of in vivo experiments designed to evaluate the hydrodynamic theory, Brannstrom and Astrom observed that an elevation in temperature 30€C above ambient failed to elicit a painful response; whereas pain was invariably elicited when there was a drop in temperature. Pain elicited from prolonged application of heat is generally of a dull nature and normally took longer to develop, in contrast to the immediate sharp pain elicited from a cold stimulus. Several investigators have suggested that the delay in response may be due to the larger volume of dentin that must be heated before sufficient movement of tubular contents can occur. It is also possible that a specific pulpal temperature must be reached before pain is experienced, and this may account for delay in response.
According to Narhi, activation of both intradental A- and C-fibers may contribute to two different types of pain sensation following heat stimulation. The pattern of nerve response appears to be a rapid A- unit action followed by a delayed C-fiber firing. The clinical features constitute an initial sharp pain following heat application and a subsequent dull pain, provided the stimulation is continued and the temperature of the pulp is elevated to about 44€C. Rapid cooling also can induce fluid flow and cause activation of intradental A-fibers. Cooling also may activate A-fibers directly, while C-fibers may respond to cold once the stimulus has reached the pulp. Activation of C-fibers, which may contribute to the dull pain induced by intense thermal stimulation of the tooth, however, appears to be associated with pulpal inflammation.
More recently Kim postulated two mechanisms whereby thermal stimuli elicit a painful response: (1) based on the hydrodynamic theory, thermal stimuli evoke dentinal sensitivity by changing the physical properties of the dentin, namely, tubular radius and dentin fluid viscosity; and (2) thermal stimuli alter pulpal microcirculation which in turn causes sensory nerve excitation by increased tissue pressure.
These studies, therefore, support Brannstrom's hypothesis of a hydrodynamic mechanism. Arguably the most significant fact from the Brannstrom and Astrom study was that pain was caused by the rapid displacement of tubular contents, and not the slow outward movement of fluid that normally occurred.
According to Narhi however, as long as the fluid flow in dentinal tubules cannot be measured in vivo, the evidence supporting the hydrodynamic theory remains unsubstantiated. Other investigators, while acknowledging that the vast amount of experimental data from both in vitro and in vivo studies appear to support the concept of a hydrodynamic mechanism of stimulus transmission across dentin, nevertheless have suggested that alternative mechanisms also may be responsible.
Recently Vongsavan and Matthews demonstrated, possibly for the first time (in vivo), that the ferocity with which fluid flows outward through exposed dentin can be sufficient to substantially reduce the inward diffusion of substances into the tubules. According to these investigators, it is also possible that excitation of intradental nerves by mechanical stimulation of exposed dentin may be due to a sudden interruption of an existing outward flow in the dentinal tubules. These results would, therefore, appear to substantiate the concept of a hydrodynamic mechanism.
Several investigators have used a neurophysiological model to evaluate dentin sensitivity. The results from these studies, together with the work of Kim and other Theory) investigators, would suggest that application of various chemical solutions (in particular K+-containing compounds) to dentin resulted in raising the intratubular K+ content which in turn rendered the intradental nerves less excitable to further stimuli by depolarizing the nerve fiber(s) membrane.
On the basis of these studies, Kim and Markowitz and Kim proposed an alternative mechanism, namely desensitization of dentin by blocking nerve activity (direct ionic diffusion).This recent hypothesis has, however, been criticized. According to Sena, Kim's work was based on deep-cut cavity preparations, with only a very thin slice of dentin between the exposed dentin surface and the pulp. In consequence K+ had only a short distance to traverse the length of the tubule. In the normal clinical situation, however, the incoming K+ (i.e., if applied in a toothpaste on exposed cervical dentin) would have to overcome the opposing pulpal pressure that produces an outward flow of dentinal fluid. Such an outward flow can prevent the inward diffusion of substances from the oral cavity. While the alternative or modified hypothesis of stimulus transmission across dentin as proposed by Kim and Markowitz and Kim appears to be an attractive alternative to the hydrodynamic theory, this hypothesis nevertheless requires further investigation.
Although observations from early histological studies appeared to indicate that dentin was directly innervated, it is now accepted that its coronal two-thirds is largely devoid of nerve fibers apart from the predentin and its most pulpal aspects. Holland, however, has suggested that this may not be correct in the light of recent immunohistochemical findings and improvements in fixation techniques which have demonstrated structural evidence of odontoblastic processes in peripheral dentin. The possibility that odontoblastic process may be of varying lengths and in some sites occupy the full extent of the dentinal tubule also has been proposed. Takahashi however, doubted whether these new approaches to resolve this problem (e.g., the extent of the odontoblastic process in dentin) had provided any further information to that previously understood. Electrophysiological recordings in experimental animals also have indicated that intradental nerve fibers are responsible for the sensitivity of dentin, and that the endings of the responding fibers are located in the pulp border area. Investigators also have failed to demonstrate any morphological evidence of a synaptic relationship between odontoblast and sensory nerve endings. Results from various animal and human studies have indicated that dentin sensitivity persists despite damage, disruption, or destruction of the odontoblast layer, which would appear to contradict the hypothesis that the odontoblast acts as a dentinal receptor mechanism. If such a synaptic relationship existed, then the presence of a neurotransmitter substance such as acetylcholine would have to be demonstrated by the evidence of acetylcholinesterase activity in the dentin. No evidence has been provided to suggest that these impulses are mediated by cholinergic activity.
While the exact mode of transmission of stimuli across dentin is still unclear, of the various mechanisms reviewed, the hydrodynamic theory appears to be the most commonly accepted. The earlier studies, such as those of Anderson and Brannstrom, were essentially physiological, concerning coronal dentin rather than studies of root dentin per se. Anderson and Matthews also cautioned against extrapolating their observations on non-carious, healthy coronal dentin to sensitive root dentin.
The earlier investigations of Anderson and co-workers reported that the ability of various chemical solutions (e.g., dextrose, CaCl2, NH4Cl, etc.) to cause pain in vivo appeared to be related to their osmotic pressure. Horiuchi and Matthews, however, reported that fluid movements caused by solutions of different chemical substances could not always be accurately predicted by their osmotic pressure alone. Horiuchi and Matthews also applied chemical stimuli to dentinal cavities of varying depths in the cat and concluded that changes in the ionic environment around the nerve endings, rather than osmotic pressures of the applied solutions, were responsible for the induction of nerve impulses. According to Matthews, the failure of 6 molal CaCl2 to excite nerves which responded by cooling (in the dog) suggested that the fibers were not excited by an outward movement of fluid through the dentinal tubules. Both stimuli have previously been shown to cause an outward movement of tubule contents in vitro. Horiuchi and Matthews also reported that application of water (at tooth temperature) to human dentin in vivo failed to cause pain, but did cause inward fluid movement in vitro. These studies would, therefore, appear to suggest that other mechanisms of pulpal nerve activation (i.e., apart from a hydrodynamic mechanism) were responsible for the transmission of stimuli across dentin. Kim suggested that identifying open dentinal tubules as the cause of dentin sensitivity may be premature, and several other investigators also have suggested, on the basis of conflicting responses to chemical stimuli that there may be more than one mechanism involved. A modification of the hydrodynamic theory, based on a neurophysiological model, has been proposed by Kim. While the concept of dentin desensitization by blocking nerve activity (direct ionic diffusion) appears an attractive alternative to the hydrodynamic theory, this hypothesis requires further investigation. Recent investigations (in the cat) by Vongsavan and Matthews, however, appear to provide evidence substantiating the hydrodynamic theory.