Summary.

A pharmacological concept of topical peripheral pain therapy with a combination of a sodium channel blocker from the class of local anesthetics and a NSAID from the class of nonselective/partially selective COX inhibitors is described. The dual mechanism of action of these substances on the cellular level interferes with pain components of both inflammatory and neuronal genesis and is functionally synergistic. This enables the combination of these agents to produce a more long-lasting analgesia than is attainable with either of the substances alone. The therapeutic indication is the symptomatic reduction of neuromuscular pain of traumatic or metabolic origin. The already more favorable safety profile for topical applications can be improved further by a possible dose reduction of the two substances.

1. Introduction

Peripheral neuromuscular pain mainly results from physical, chemical, or biological-infectious noxae with tissue damage in the musculoskeletal system. It therefore affects the functionally involved muscles, tendons, and joints as well as the structurally or functionally involved bones, ligaments, and cartilaginous structures. But since this always involves an interaction between neuronal and biochemical-muscular components, we prefer the term neuromuscular pain to myalgia or musculoskeletal pain.

The pain symptoms as they occur, for example, with so-called nonspecific muscular tension, after blunt muscular trauma, with joint trouble, and also after incisions during operations, are produced by a local cascade of biochemical reactions in the tissues. Regardless of the nature of the causative stimuli, these always lead to an irritation of local afferent nerve functions in the affected area via the release of algogenic mediators, disturbances of capillary permeability, and local edema. In the course of the initial local pain process, the reactive central pain perception and its emotional assessment via the limbic system first seem to be of less pathological importance than during the transition to chronic pain. Pharmacologically, these pain symptoms are presently being treated mainly with systemic antiphlogistic agents, and with chronic conditions increasingly with central psychoactive drugs sometimes together with psychosomatic or physiotherapeutic measures.

2. Present Status of Pharmacological Therapy

At present, non-steroidal anti-inflammatory drugs (NSAID ), whose broad spectrum includes well-known substances like acetylsalicylic acid, diclofenac, and ibuprofen, are used as the main pharmacological principle for alleviating peripheral neuromuscular pain symptoms and painfully impaired locomotor function. But NSAIDs are primarily antiphlogistic agents that have a secondary analgesic action only via their anti-inflammatory effects. It can be concluded from their widespread use that predominantly inflammatory processes are assumed to be the main cause of peripheral neuromuscular pain symptoms. But all systemic NSAIDs show a significant rate of undesirable effects, especially gastrointestinal disturbances. In addition, prolonged use of systemic NSAIDs can cause liver and kidney failure. The recent advances in NSAID analgesia have focused primarily on reducing gastrointestinal side effects. Clinically, the newer specific inhibitors of cyclooxygenase-2 (COX-2) show a lower overall rate of gastrointestinal side effects. But since COX-2 has been detected in the CNS, stomach, and kidneys as a constitutive enzyme, the therapeutic expectations with COX-2 selective inhibitors (e.g. that also renal side effects can be avoided) have not been confirmed to the extent hoped for. In addition, since the healing of existing lesions of the gastric mucosa is mainly the result of the prostaglandins produced by COX-2 (1), it was also not unexpected that even specific COX-2 inhibitors may retard the healing of an existing gastric ulcer (2). Within the context of the recent conditioned withdrawal of the widely used COX-2 inhibitor rofecoxib from the market, this was justified with the establishment of an increased risk of heart attack and stroke. Although this aspect of the safety profile of COX-2 inhibitors was known, it was not previously considered to be of primary importance. The extent to which this effect is substance-specific or involves a risk for all COX-2 specific inhibitors has not yet been established.

For transcutaneous forms of administration with fewer side effects (3), whether and to what extent topical NSAIDs have analgesic action is the subject of controversy. In an analysis of the therapeutic results of 49 placebo-controlled clinical studies (4), 37 of the studies were included in the analysis of patients with acute pain (tissue traumas, strains, sprains). The end point was an at least 50% reduction of pain after one week of treatment. Here, topical NSAIDs were significantly more efficacious than placebo in 27 of 37 (72.9%) of these studies. Compared with placebo, 1.7 times more patients from the NSAID groups attained the end point; however, an average of about 4 patients had to be treated in the NSAID group to achieve one successful therapeutic result. In the other 12 studies, the efficacy of topical NSAIDs for chronic pain, above all arthritis, was analyzed. The end point here was an at least 50% reduction of pain after a 2-week treatment. In 7 of these 12 studies (58%), the topical NSAIDs were significantly more efficacious than placebo. Other more recent meta-analyses (5,6) have also confirmed the usefulness of topical NSAIDs in the treatment of acute neuromuscular pain, but also that the long-term effectiveness is even less substantiated here (7). Thus, although it seems unquestionable that topical NSAID applications show a significant analgesic effect compared with placebo, the efficacy percentage rate is not convincing. It seems to be even lower in the area of pain with only mild inflammatory components, which often occur with chronic pain conditions. Therefore, these findings also show that a differentiated course of action with regard to the presence of an indication suitable for this substance group, especially the existence of an inflammatory process, must also be taken into account for topical NSAID applications.

Since transcutaneous NSAID formulations show more pronounced effects when they are applied to open skin sites or large skin areas, it seems natural that weaker effects are commonly attributed to pharmacokinetic factors (i.e. weaker effect is primarily due only to a lower skin absorption). However, this would therefore be interpreted only quantitatively, as a dose-effect problem. With regard to the above-mentioned analgesic effects of the NSAIDs in clinical studies, such an assumption about the reason for the decreased effect must also consider a pharmacodynamic cause, which again raises the question of the extent to which the topical administration followed a determination of an indication suitable for NSAIDs.
The NSAIDs affect only some of the series of sequential interactions of the cascade of the cellular pain mechanism that also have pain relieving effects, specifically those which are associated with or a consequence of tissue inflammatory processes. In this way, the peripheral mechanism of the NSAIDs causes primarily only an inhibition of the cyclooxygenases, which mainly reduces the formation of prostaglandins. Without going more specifically into detail about the wide spectrum of activity of the prostaglandins, it is known that they play a role in the development of inflammatory processes and fever and increase the sensitivity of nociceptors. The basic mechanism in the inflammatory response which stimulates nociceptors directly and sensitizes them to further stimuli is the release of various mediators from damaged tissues. It includes amongst others bradykinin, histamine, leukotrienes, serotonin, hydrogen and potassium ions, substance P, and peptides from the sensory nerve endings themselves. In addition, prostaglandins also exert a pain-relevant effect on the contraction of the smooth musculature.
On the other hand, the NSAIDs exert no direct effects on the subsequent process of neuronal transmission of pain information which is important for the pain condition. In contrast to inflammatory pain, COX-2 presumably also plays no direct role in nociception, which is shown by the only slight effects of the NSAIDs in some classical pharmacological models of nociception (8). That the production of peripheral analgesia involves additional mechanisms is further emphasized by the fact that still no direct correlative relationships have been found for the NSAIDs between their antiphlogistic and antinociceptive effect components.
All in all, it can be concluded here that frequent unsatisfactory clinical therapeutic results with topical transcutaneous NSAID application for the treatment of peripheral neuromuscular pain are observed when local inflammatory components, and their pathological sequelae or accompanying symptoms, are not present, or when these become gradually less pronounced during the treatment phase. But if the local pain process can only be influenced primarily on the subsequent level of disturbed neuronal processes, it is also largely unaffected by the biochemically-metabolically oriented spectrum of NSAID activity. In such cases, a therapeutic failure cannot be attributed to the ineffectiveness of the NSAID active agent or to its to low local concentration in the sense of a dose-effect relationship. On the contrary, a pharmacodynamic problem is then involved: the pain condition was treated with a qualitatively unsuitable substance. NSAIDs are primarily antiphlogistic agents and therefore only exert a secondary analgesic effect if an inflammation is also present.

As an additional concept in the treatment of peripheral pain symptoms with predominantly neurogenic components, target-specific injections with local anesthetics (LA) have proven to be clinically successful. This long-established method is partly based on experiences with so-called neural therapy (“wheal therapy”, local infiltration with procaine) as well as findings on so-called trigger points (9). In the early historical phases, its aim was to block the action of afferent nerves or nerve groups via its anesthesia in order to suppress pathological (including vegetative) local reflexes. As another differentiated principle here, a noninvasive administration of LA by means of topical transcutaneous plaster systems which we introduced into the field of the treatment of neuromuscular pain has also proven its value in recent years both as a topical application with back pain and muscle tension as well as in the area of the topical treatment of headache (10,11) In this connection, it was also shown that local anesthetics cannot only be viewed and clinically employed as conventional anesthetics but also as specific neurally applied analgesics, as specific antineuralgic agents. The usefulness of this therapeutic approach was confirmed in clinical studies on, for example, neuropathic pain and postherpetic neuralgia (12,13). For the indication of postherpetic neuralgia, a topical analgesic treatment with plasters containing 5% lidocaine is also authorized by the US FDA. This noninvasive pharmacological concept has also resulted in other treatments, e.g. in the area of post-incisional pain after hernia operations, which have also been shown to be efficacious in clinical studies (14).
Chemical amide- or ester-type local anesthetics, e.g. amide-type lidocaine, show as a mechanism of action an inhibition of the rapid Na+ ion influx into nerve fibers; that is, they act as sodium channel blockers (SCB). Via this specific effect, they block the impulse conduction of all regional nerve fibers (15) in their field of administration. However, since the pain-conducting fibers in the cutaneous sensory network (i.e. the myelinated fibers and unmyelinated C fibers) are anatomically clearly thinner than the motor fibers, and also thinner than the fibers responsible for touch and pressure and the proprioceptive fibers, different topical effects, related to the lower LA dose required, can be distinguished. This was also specifically investigated for lidocaine (16). Another essential aspect for this transcutaneous topical approach with LAs was to conceive of pain signals as irregular neuronal information patterns to the CNS which can be regularized via lower LA doses. These predictions of an effect on „ectopic“ nerve impulses was experimentally verified for lidocaine in recent years (17,18). Since effects on the electrical potential of the nerve fiber cells can therefore be already produced with doses below those that have local anesthetic efficacy, a peripheral reduction of pain requires significantly lower LA doses than for a complete inactivation of neuronal transmission or anesthesia. In addition to the anatomically thinner structure of sensory pain-conducting fibers, which is advantageous in terms of the dosage for a specific analgesia, this also results in the therapeutic possibility of an “analgesia without anesthesia”. With regard to their mechanism of action, the SCBs from the class of LAs therefore function as pain therapeutic agents like a kind of “neural firewall”. Therefore, a transcutaneous analgesia can also be distinguished by its selectivity from a transcutaneous anesthesia, whose inactivation of motor functions is undesirable in the case of the therapeutic aim of only analgesia. Since LA-type SCBs reduce the extent of the influx of pathological signals into the level of the CNS perception and interpretation via such an effect, abnormal systemic reactions, for example those via inadequate activation of the autonomic nervous system, can be lessened in this way. But on the other hand, these effects of the LA-type SCBs apply specifically only to this part of the neuronal level. Unlike the NSAIDs, they neither exert direct pharmacological effects on the biochemical inflammatory processes in the tissue (15) nor do they have antipyretic effects.

3. Model of a Cellular Mechanism for a Functional Pharmacodynamic Synergy

Therapeutically, the conceptual aim of the use of a combination of an NSAID and sodium channel blocker-LA is to combat the interactive cellular pathophysiological pain processes in a pharmacologically interactive way. By the concurrent use of these active substances, both the initial pain-inducing biochemical effects (inflammatory-edematous) and their sequelae on the afferent sensory conduction of the pain-related information are inhibited in the affected tissue.
As discussed at the start, the analgesic mechanisms of action of the NSAIDs on the one hand and the LA-type SCBs on the other exert their pharmacological effects only at different points in the peripheral cellular pain cascade. This is summarized in Figure 1. The NSAIDs reduce via COX inhibition primarily the pain-inducing processes of cellular inflammation (i.e. local release of algesic substances and concomitant capillary effects with edema formation). They therefore play their mainly analgesic role by inhibiting such factors that are involved in the initiation of pain associated with local inflammatory processes. But these processes also exert pathophysiological effects on the local nerve functions. Consequently, the tissue-relieving anti-edematous effects of the NSAIDs also lead to decreased mechanical stimulation of the local sensory system. The analgesic effect can therefore also be regarded as a consequence of the reduction of such local neuronal stimuli.
Unlike the NSAIDs, the local anesthetic-type sodium channel blockers exert no direct effects on the level of the inflammatory process. The only exert analgesic effects on the reactive neuronal processes of pain conduction, particularly the levels of transformation and conduction, and indirectly on the subsequent processes of central pain perception. The latter results from a reduction of pathological information to the CNS, which with the lower LA dosages primarily concerns the A fibers and C fibers as an anatomically more sensitive substrate. Functionally, the transmission of nociceptive information from the damaged tissue to the interface to the central neurons is inhibited here; that is, before they can be activated. Thus, such a suppression of the transmission of pain information can be reasonably regarded as a “firewall effect”.

Figure 1: Dual mechanism of action of SCB-LA and NSAID in the cellular local pain process

Therefore, each of the substances applied fill an effect gap left by the other substance class. This spectrum of action is not possible with either of the individual substances alone (Table 1). Furthermore, since the effects are exerted via topographically different receptors, interactions which are manifested as functional pharmacodynamic synergism in the sense of addition or potentiation are also to be expected (Table 1). Thus, the dual action of the combination produces a therapeutic effect potentiation and, therefore, a sustained reduction of the pain symptoms, or conversely makes it possible to reduce the doses of the components.

Table 1: Summary of relevant pain reduction effects of SCB-LA and NSAID and their synergistic effect on the cellular level

Another practical advantageous effect of the combination is that the anti-edematous effect of the NSAID improves the local pharmacokinetic effect conditions for the SCB-LA. For example, the transcutaneous effect of SCB-LA depends on the pH value of the tissue. However, due to the edema-induced increased anaerobic glycolysis (local lactacidosis), inflamed tissue shows a lower pH value than normal tissue. The SBC-LAs are less efficacious in inflamed regions because here their chemical balance between protonated and non-protonated form is shifted in favor of the protonated component, which in turn lowers its penetration capability(20). This effect inhibition is reduced by the antiphlogistic NSAID effect.

4. Safety and Toxicity Aspects

The topographically different sites of action of the active agents opens up the possibility of additive effects with a possible dose reduction, and therefore an improved safety potential. For both classes of active substances, extensive clinical experiences on the topical cutaneous tolerance are already available. Since SCB-LAs (e.g. lidocaine) are already mainly metabolized during the course of their retarded transcutaneous absorption, a systemic toxic effect level is avoided (12, 13, 21). To indicate a respective safety margin: To achieve a toxic plasma level for lidocaine, which is estimated at 5 ug/ml serum, it is necessary to inject 200 mg lidocaine as an intravenous bolus or 1000 mg subcutaneously.
Plasma concentrations after topical administration of NSAIDs are low and produce plasma concentrations 5% or less than the maximum oral concentration. In clinical studies on the transdermal administration of the NSAID diclofenac in plaster form, a good tolerance was reported (22). And in a clinical study with such NSAID plasters for the treatment of blunt traumas, the mild local skin symptoms of pruritus and rash were reported as the most common side effects with about the same frequency as in the placebo group (23). At present, there are no known kinetic interactions between NSAIDs and SCB-LAs, and due to their differing receptor affinity, they are also not to be expected. All in all, topical transcutaneous treatment can be regarded as safer than systemic treatment from the general pharmacological point of view (24). With regard to administration with post-incisional pain, an antibacterial effect of the lidocaine should also be pointed out (25, 26), and also the fact that LAs may obviously have advantageous effects on wound healing or least do not disturb the wound healing process (27, 28).

5. Conclusions

It can be deduced from the mechanism of the cellular pain cascade that an interaction of inflammatory processes with a reactive neuronal transmission disturbance exists. Here, one of these processes can at times be regarded as the main pain-inducing component. In the course of the process from a causative tissue noxa to the completion of tissue regeneration, a dynamic transition from a primarily edematous-inflammatory phase to one with primarily only neuronally accentuated characteristics takes place on the cellular level. This process may also sometimes recur in phases (e.g. via the reactivation of an inflammatory process). The current extent of the respective components then also has a direct effect on the therapeutic possibilities. With acute pain conditions, for example, the pharmacokinetic effect conditions for local anesthetics to reach the C fibers may be worsened due to the primary inflammatory-edematous components (local acidosis, diffusion). This also applies to the special case of neuropathies with the formation of perineural edemas. On the other hand, chronic peripheral pain tends often to show a predominance of neuronal irritation with even less pronounced signs of inflammation. However, topical NSAIDs can exert only slight effects on the process of neuronal transmission.

In therapeutic practice with regard to peripheral neuromuscular pain, it presently seems often rather common to equate the pain primarily only with an inflammatory process, since a determination of the acute primary main components seems to be impossible with the usual clinical measures. This seems especially to be the case within the context of the treatment of so-called muscular rheumatism for which the topical administration of purely antiphlogistic formulations is particularly common. In fact, it was already found in older studies on so-called myogelosis that inflammatory components are histologically only very weakly discernable. For example, optical microscopic examination of biopsies from painful myogelosis showed no indication of the presence of an inflammation. Only with electron microscopic resolution were changes in the myofilaments and in the endothelial cells of the muscle capillaries detected (29, 30, 31). This was attributed to relative hypoxia of the muscle cells which occurs with continuous tension of isolated muscle segments (which clinically corresponds to „myogeloses“). Reported is also a two-stage process; an initial neuromuscular dysfunctional stage, and a subsequent dystrophic pathological stage. The primary cause of this excessive tonus here is therefore a neuronal irritation, and it can be produced as such by, for example, thermal stimuli or functional false postures. Therapeutically, this is more easily treatable on the nociceptive neuronal level. In this case, this also explains a relative high number of unsatisfactory analgesic effects of topical NSAID monotherapy compared with a primary neuronal approach with SCB-LA. Here, an inhibitory effect of the NSAIDs on prostaglandin synthesis no longer makes sense. Furthermore, it seems to be another approach to explaining why it has not yet been possible with NSAIDs to find a direct correlation between antinociceptive and antiphlogistic effects. On the other hand, by means of a simultaneous attempt to treat both the neuronal nociceptive and inflammation-induced pain by means of a combination of NSAID and SCB-LA, both pain components can be combated here.
Furthermore, since the effects of NSAIDs und SCB-LAs are exerted via topographically different receptors, interactions which are manifested as functional pharmacodynamic synergism in the sense of addition or potentiation are also to be expected (see Table 1). Thus, the dual action of the combination produces a therapeutic potentiation of the effect and, therefore, a sustained reduction of the pain symptoms, or conversely makes it possible to reduce the doses of the components. First preliminary results from an experimental clinical phase I pilot study (double blind, randomized, placebo controlled, 10 healthy volunteers) support this assumption (32). In this experimental model measurement of pain perception following a defined freeze lesion of the thigh was performed by means of von Frey hairs. A combination with two 25% dosages of both the amide-type SCB-LA lidocaine and the NSAID diclofenac reached higher efficacy compared to the 100% dosages of each of the single drugs.

Figure 2. Topical pain reducing effects on a freeze induced skin lesion. Comparison of the NSAID diclofenac (Diclo 100) and the SCB-LA Lidocaine (Lido 100) with a combination that contains 25% dosages of each agent (Lido 25 + Diclo 25). The AUC values represent the cumulative von Frey Hair weights required in the test period to induce a pain sensation .

The AUC effect for the different dosages can be summarized as . Such a relation indicates pharmacologically a synergistic effect of the combination. However, with respect to a low sensitivity and high variability of the von Frey hair method, in particular in a small cutaneous lesion and the small number, this didn’t yet reach statistical significance. The results in such experimental models are also dependent from the respective grade of the neuronal and/or inflammatory pain component, thus from the employed noxious stimuli. For example it cannot be exluded that a freeze lesion also damages C-Fiber functions, which then can reduce effects of an SCB. Further investigation with different other models are underway.
Finally, and in this context, it may also be possible to extend the general pain definition in that sense that peripheral pain appears a mechanism that bases on a coexisting dynamic relation of two components, a terminal nerve irritation and a tissue damage with inflammation.

References [1] Gretzer, B., Ehrlich K., Maricic N., et al., Br. J. Pharmacol, 123, 927 (1998) [2] Brzozowski, T., Konturek P.C., Konturek, S.J., et al, Microsc. Res. Tech. 53 , 343 (2001) [3] Karzel, K., Liedtke, R. K., Arzneim. Forsch./Drug Res. 39 (II), 1487 (1989) [4] Moore, A., Tramer M.R., Carroll D., et al., BMJ 316, 333 (1998) [5] Mason, L., Moore R.A., Edwards J.E., et al., BMC Musculoskelet. Disord. 5 (1); 28 (2004) [6] Mason, L.,BMC Family Practice; 5, 10 (2004) [7] Lin, J. Zhang W, Jones A, et al, BMJ 329, 324 (2004) [8] Santos, A.R., Vedana E.M., De Freitas G.A., Inflamm. Res. 47 (7), 302 (1998) [9] Travell, J. T., Simons, D. G., Myofascial Pain and Dysfunction, Vol. I, pp. 5-44, Williams & Wilkens, Baltimore, 1983 [10] US Patent 5,776,952 [11] US Patent 5,840,755 [12] Galer B. S., Jensen M.P., Ma T., et al., Clin. J. Pain. 18 (5), 297 (2002t) [13] Argoff, C.E., Clin. J. Pain. 16 (2 Suppl) S62 (2000) [14] unpublished clinical data, EpiCept Corporation (2004) [15] Strichartz, G. R. (Edit.): Local Anesthetics, Handbook of Experimental Pharmacology, Vol. 81, Springer, Berlin-New York, 1987 [16] Sakai, T., Tomiyasu S, Yamada H., et al., Anesth. Analg. 98 (1): 248 (2004). [17] Khodorova, A., Meissner K., Leeson S., Strichartz G.R., Muscle Nerve, 24 (5), 634 (2001) [18] Persaud, N., Strichartz G.R., Pain, 99, 333 (2002) [19] Pedersen J. L., Callesen T., Moiniche S., et al., Br. J. Anaesth. 76 (6), 806 (1996) [20] Hadgraft, J., Valenta C., Int. J. Pharm. 200 (2), 243 (2000). [21] Gammaitoni, A.R., Alvarez N.A., Galer B.S., J. Clin. Pharmacol. 43 (2), 111 (2003). [22] Gallacchi, G., Frentzel A., Heidecker J. L., Schweiz. Rundsch. Med. Prax. 92, 2201 (2003) [23] Predel, H. G., Koll, R., Pabst, H., et al., Br. J. Sports Med. 38 (3), 318 (2004). [24] Karzel, K., Liedtke, R.K., Einführung in die Arzneimitteltherapie, 2. erw. Aufl. S. 104-115, “Applikationsort Haut ”, G. Fischer, Stuttgart-New York, 1985 [25] Thompson, K.D., Welykyj S., Massa M.C. et al., J. Dermatol. Surg. Oncol. 19 (3), 216 (1993) [26] Stratford A.F., Zoutman D.E., Davidson J.S., Plast. Reconstr . Surg. 110 (5), 1275 (2002) [27] Eroglu, E., Eroglu F., Agalar F. et al. Eur. J. Emerg. Med. 8 (3), 199 (2001) [28] Blanke, W., Hallern, B.V., Eur. J. Emerg. Med. 10 (3), 229, (2003) [29] Awad, E. A, Arch. Phys. Med 54, 440 (1973) [30] Fassbender H.G., Wegner K, Z. Rheumaforsch. 32, 355, (1973) [31] Fassbender H.G: Pathology of Rheumatic Diseases, Chapter 13, pp. 303-314, Springer, New York, 1975. [32] Unpublished clinical-pharmacological data; pharmed (2005)