Topical ambroxol for the treatment of neuropathic pain

By way of example, we present the case histories of seven patients with neuropathic pain that was resistant to other/standard treatment and who benefitted from topical ambroxol. Within the setting of an outpatient pain clinic they were treated in a peripheral, defined part of the body after numerous other therapeutic attempts with authorized substances had proved unsuccessful or impossible. The preparation used was a 20 % cream (ambroxol cream 20 %, 50.0 g: ambroxol 10.0 g, dimethyl sulfoxide 5.0 g, made up to 50.0 g with Linola cream), which was prepared pharmaceutically for each individual patient. In tests this formulation was shown to possess the highest concentration of still soluble ambroxol.

While test treatments were administered during the consultation, in the majority of cases the observations reported here were collected by the patients after they themselves had applied the preparation to the painful area over a prolonged period at home. Clinical observations and reports were documented in the medical records and retrospectively served as the basis for the case reports.

All the patients gave their written informed consent both to the individual test treatment with ambroxol-containing creams (which were prepared individually and are not available as proprietary preparations) and also to the use of the anonymized reports, findings and possible images/videos. The observations presented here by way of example refer to a period of 4 years overall and were documented and reported on the basis of the German professional code of conduct for doctors (particularly §15) and the requirements of the Declaration of Helsinki (§37). The treatments were prescribed outside the health insurance reimbursement scheme.

Dynamic allodynia was present in six of the seven patients and a hyperalgesia in response to pinprick stimuli in four patients. Both forms were present in four cases, while one patient suffered from neither. Dynamic allodynia on its own was observed in two cases, and no patient was sensitive exclusively to pinprick stimuli (Table 1). Irrespective of these configurations, all patients were sensitive to ambroxol.

Six of the seven patients had previously used lidocaine plasters, which proved ineffective in four cases and helpful in two patients, one of whom was unable to tolerate them. Both of the patients experiencing pain relief also reported this with ambroxol. By contrast however, all four patients in whom the lidocaine had proved ineffective profited additionally from the effect of ambroxol.

Three of the seven patients had also received capsaicin 8 % as topical treatment. In two patients this was helpful, and one continues to use the medication. Both patients whose pain was relieved by capsaicin 8 % also experienced this relief with ambroxol. By contrast however, the patient in whom the capsaicin had proved ineffective benefitted from topical ambroxol.

Born in 1942, this man presented in October 2010 (clinically) with neuropathic pain in both feet. He felt as if he was running on coals, while his right foot felt as if it were gripped in a vice. Walking or gardening was almost impossible. The foot pain could not be explained in orthopaedic terms, either on the basis of radiographs or as a result of a spondylodesis of L4/5 in 2007, and no polyneuropathy was detected on neurographic investigation. The problem was thought to be ‘likely of vertebral origin’. Clinical examination showed pronounced dynamic allodynia and hyperalgesia in response to pinprick stimuli in the arch of the right foot, and to a lesser extent in the left foot. Gabapentin had been prescribed up to a dosage of 3 × 600 mg, supplemented by buprenorphine 20 µg/h, and no further dose increase were possible in either case. Topical lidocaine 5 % plasters had not proved helpful.

The first test treatment with topical ambroxol was started in June 2011, with a very successful outcome: within 5 min the stabbing pain, with an initial intensity of 8/10 on the NRS scale during the heel-to-toe roll of the right foot, and the touch sensitivity of the forefoot disappeared completely for over 8 h. The intermittent treatment with topical ambroxol has been continued to date successfully (Fig. 1). The pain has also meanwhile been treated successfully on 11 occasions with capsaicin 8 % plasters, and the patient now uses ambroxol cream in phases when the pain recurs and the capsaicin effect is wearing off prior to the next application. He now takes gabapentin only in the evening at a dosage of 300 mg, while the opiate has been discontinued. Walking and gardening are now possible again. No skin reactions or other side effects have occurred during the treatment that has proved effective till now, without any changes, for 4 years. This case report has been documented repeatedly on video.

This patient presented with unusual pains: he complained of extremely painful cold sensations in both previously amputated phantom feet. A below-knee amputation of the left leg had been required for pAOD and diabetes mellitus in 2008, followed by the right leg in 2009. The cold pains, with an intensity of NRS 7–9/10, occurred sporadically several times a week, lasted from a few minutes to many hours and frequently woke the patient at night. He described the cold sensation as shifting between the toes and balls of the feet and said that heat applied to the stump was of limited benefit. The cold phantom pain could be triggered by a cold environment, but also by visually perceived cold (e.g. images of snow). He had not experienced the cold sensation prior to amputation (in terms of preamputation recall), and the stump itself was usually warm. Both stumps showed slightly increased cold perception on the distal lateral sides, on the right with additional dynamic allodynia (without pinprick hyperalgesia). Following many unsuccessful treatments (includes opiates and anticonvulsants), the topical ambroxol 20 % cream managed to produce a relevant effect: the extreme cold sensation in the phantom regressed distinctly after approximately 15 min for several hours, and the phantom limb felt warmer. This effect has persisted unchanged for over 11 months now. No skin changes of any kind or any other side effects occurred on the stump, and the case report is documented on video.

This 58-year-old female patient received a total knee replacement in November 2010, after which considerable pains persisted. Clinical examination revealed an extensive dynamic allodynia and pinprick hyperalgesia medially as a sign of central sensitization. There were no signs of inflammation or movement-related pain. The existing treatment with buprenorphine 10 ug/h had been replaced by tapentadol because of the neuropathic nature of the pain, but this had proved no better in relieving the pain. After capsaicin 8 % plasters, the pain became slightly less frequent and of shorter duration. Since the knee pain was still persisting substantially by April 2014 and lidocaine plasters had also not been very helpful, topical ambroxol 20 % was used during the consultation. After just 15 min the patient reported clear pain relief: the burning and stabbing had subsided distinctly, while a ‘raging feeling’ in the knee disappeared almost completely. After applying the cream repeatedly over the next few months, she observed a reduction in pain for 4–6 h after approximately 30 min in each case, from an average of NRS 8/10 before treatment to 4/10 NRS (Fig. 2), and occasionally even down to NRS 1/10. With the most extreme pain (NRS 10/10), she found that the pain no longer returned to this level after the treatment. She has been applying topical ambroxol 20 % cream regularly for 11 months now, without any side effects or skin reactions. The case report is documented on video.

The 38-year-old patient presented in April 2014 with deafferentation pains in the left arm after a plexus lesion (motorcycle accident in 1997), with subsequent mechanical allodynia in hand and forearm. A nerve graft proved unsuccessful, as did ketamine, gabapentin and a lidocaine infusion. Mirror therapy ended with a worsening of the pain for several days, while amitriptyline produced excessive sedation. Although cannabis (in connection with a clinical study) managed to reduce the pain by 60 %, it also resulted in substantial mental impairment. His medication now consisted of pregabalin 2 × 300 mg and duloxetine 60 mg. Slight innervation of the biceps muscle and shoulder elevation were possible, as was mental motor imagery to modulate the phantom pain. The patient experienced three types of pain in the arm: a ‘burning pain’ (started by the application of cold to the hand), a ‘crushing underlying pain’ (with no trigger) and ‘shooting tingling pains’. The pain intensity ranged from NRS 4–8/10, and there was no dynamic allodynia at the elbow. Sensations could be triggered in the phantom arm by cold stimuli to the subclavicular area and, as referred pains, from trigger points from the subclavian and pectoral muscles. Lidocaine plasters merely changed the nature of the pain, while trigger point treatments and tapentadol were not tolerated. A test was therefore initiated with topical ambroxol 20 % cream over the pectoral muscle, which successfully relieved the shooting and tingling pains in most cases from NRS 8/10 to 4/10. Since then the effect has been described as starting after approximately 15 min and persisting for 4–6 h, which proved crucially important particularly in falling asleep. However, the ‘deep underlying pain’ remained unaffected. The patient was especially happy about the spontaneous cessation, without triggers, of repeatedly occurring spasms and cramps, which enabled him to continue everyday tasks at these times. The patient’s descriptions are documented on video.

This 55-year-old patient presented in August 2008 with a postherpetic neuralgia on the right side of the chest (intensity 5/10) that had been present for 2 months. Step I and Step II analgesics only reduced the pain slightly, while lidocaine plasters proved very effective. Clinical examination revealed an almost circular, dynamic allodynia at T6-8 on the right. He was initially treated successfully with increasing doses of gabapentin, and his night-time sleep improved with amitriptyline drops. Since other sedating substances were not possible and capsaicin 8 % remained unsuccessful, the use of topical ambroxol 20 % cream was tried in April 2012. Although this was unable to control the pain adequately on its own, it did manage to reduce the pain, starting after 30 min, from NRS 6/10 to 4/10 (Fig. 3) for a period of 4–6 h. The patient applies the cream for pain attacks and for those areas that are not well covered by the lidocaine plasters, which he continues to use. No adverse effects or skin reactions have been reported after what is now more than 3 years of use.

This patient presented in November 2013 with neuropathic pain in the arch of the left foot and multiplex neuritis as a result of vasculitis. He described two types of pain: a permanent underlying pain that interfered considerably with this everyday activities, and he also reported attacks of fierce pain, particularly in the evenings while watching television and at night, with an intensity up to NRS 8/10. Meanwhile he has been receiving treatment for depression with duloxetine 60 mg and lithium. Lidocaine plasters, peripheral analgesics and low-dose tilidine had already been tried without success. His night-time sleep was now deepened with amitriptyline drops, while the dosage of the tilidine was cautiously increased. Since the central nervous tolerability threshold had already been reached at the second visit in December 2013, a test was arranged with topical ambroxol. This proved highly effective: the patient reported a reduction in his evening, neuropathic pain from NRS 6/10 to 2/10, starting after 15 min and lasting for over 6 h. He was able to watch television without interruption, while his night-time sleep (after applying the cream immediately beforehand) was usually good. The zolpidem that he had been using for a long time was now unnecessary. When he was woken by pain, further application of the cream produced rapid pain relief. After 4 months of use, the permanent underlying pain during the day disappeared almost completely. A notable finding was the absence before treatment in this patient of any kind of allodynia or hyperalgesia in response to dynamic or pinprick stimuli or to cold or draughts of air. Having applied the cream for a period of 17 months now, the patient continues to use this treatment (Fig. 4) and has not reported any form of skin reactions or side effects to date. The patient’s statements have been documented repeatedly on video.

After contracting a zoster infection of the V2 branch on the left in June 2014, this 91-year-old female complained of facial pain up to the intensity of NRS 8/10. Affective pain perception was considerable, and her night-time sleep was greatly disturbed. Diclofenac, tramadol and metamizole were not sufficient. Her night-time sleep was initially improved with a few amitriptyline drops, but an increase in the tramadol dosage failed due to constipation and sedation. Lidocaine plasters were effective, but had to be discontinued due to skin reactions. Since the dosage of gabapentin (100 mg) could only be increased to a limited extent and capsaicin 8 % did not appear to be indicated in this elderly woman, ambroxol 20 % cream was attempted. She has now been using this for 11 months and describes the effect as clearly pain-relieving, and felt that her cheek became ‘calmer’ approximately 15 min after each application. If she awoke at night, she was able to fall asleep again without pain sometime after a further application. This use has also been continued now for 10 months without any adverse effects and is documented on video.

Ambroxol is structurally similar to known local anaesthetics (‘Löfgren structure’) and, like these, binds to a specific binding site of the neuronal sodium channel [27]. It likewise blocks the sodium influx at voltage-dependent sodium channels, which leads to a reduction in the action potential frequency and thus in intraneural signal transduction.

However, certain properties suggest that ambroxol is of particular interest for the treatment of neuropathic pain. Firstly, it is very potent: ambroxol inhibits neuronal sodium channels at an approximately 40-fold lower concentration than lidocaine [47], thereby alleviating neuropathic pain after local [29] and systemic application [13]. Secondly, in animal studies (and in contrast with lidocaine), it blocks nociceptor-specific sodium channels more potently than other neuronal sodium channels [47]. This may explain why four of our patients in whom lidocaine had previously proved ineffective nevertheless profited from ambroxol, which also produced a very powerful effect in animal models of neuropathic pain after systemic administration [14, 16], at least as strong as that of gabapentin and at dosages that can certainly be employed in clinical practice [14, 49].

Voltage-gated sodium channels (Na_v; v = voltage gated) in nerve cell membranes play an important role in signal transduction, including in nociceptive neurons. Nine subtypes are described (Na_v 1.1–1.9) [8]. The neuronally expressed channels Na_v 1.1, 1.2, 1.3, 1.6 and 1.7 are inhibited with high affinity by tetrodotoxin (TTX), the toxin of the pufferfish. The subtypes expressed predominantly in nociceptive neurons, that is Na_v 1.8 and 1.9, are TTX-resistant [37]. Ambroxol blocks TTX-resistant sodium channels to a much greater degree than TTX-sensitive channels. The TTX-r-subtype Na_v 1.8, which is preferentially blocked by ambroxol, is expressed particularly in spinal ganglion cells [2]. In nerve injuries, it appears to play an important role in the sensitization of nociceptors and thus contribute to the development and maintenance of neuropathic pain [3, 21, 26, 36]. It is thought to be particularly important in pain caused by cold [52]. A connection with the observation of ‘warming of the ice-cold phantom leg’ by the application of ambroxol to the stump of patient 2 remains to be seen. Na_v 1.8 blockade also proved to be clearly analgesic in animal studies [11, 20]. So the improvements in our patients are credible not just on the basis of its continued use for prolonged periods.

After nerve injuries, the redistribution or expression of sodium channels can occur in primary nociceptive neurons, with a resulting low threshold for triggering pain or for the hyperexcitability of DRG neurons. The subtype Na_v 1.8 is selectively expressed in nociceptive, sensory neurons [2, 6, 35, 52], but to a greater extent in nerve pain models [15, 44] and also in patients with neuropathic pain [24, 50]. This is also thought to be important in both animals and humans for the spontaneous activity in neuromas [5, 24, 36], which might have played a role in our amputated patients (case report 2). This may also explain the reduction in attacks of pain in all our five patients concerned. Increased Na_v 1.8 expression also occurs in diabetic polyneuropathy [30], small-fibre neuropathy [19], radiculopathies [17, 44], trigeminal neuralgia [41], but also in arthritis [39, 43] and bone metastases [33]. Na_v 1.8 blockade (e.g. produced by ambroxol) is also ultimately considered as an option for the treatment of neuropathic pain, because investigations with Na_v 1.8-free mice and Na_v 1.8-blocking substances have shown lower mechanical, thermal and visceral hyperexcitability in the animal model [1, 4, 11, 18, 20, 21, 45, 46].

Although ambroxol binds to the same local anaesthetic binding sites as lidocaine and amitriptyline, none of the substances used to date for neuropathic pain (includes local anaesthetics, antidepressants and anticonvulsants) have shown any relevant selectivity for the Na_v 1.8 channel comparable with that of ambroxol [27, 48]. The sodium-channel blockers used for treating pain, that is amitriptyline and carbamazepine, also do not block the channel types selectively [7, 42]. The selective blockade by ambroxol of the Na_v 1.8 channel type, which is not represented in the cardiac or central nervous systems, may therefore prove clinically beneficial. At least we did not receive any reports of corresponding adverse effects in any of our patients, including in those who have been using the substance topically for a long time now. This is also not surprising in that, despite the potent sodium channel blockade, the tolerability of ambroxol itself is very good after systemic administration: even intravenous doses as high as 1 g (for promoting prenatal lung maturation and the treatment of atelectasis) are well tolerated [51, 53]. There are even isolated reports of doses of up to 3 g a day being administered for up to 53 days [9, 28, 38] and the oral administration of 1.3 g ambroxol a day for 33 days [22]. After animal studies showed that high-dose ambroxol was clearly analgesic and well tolerated, Gaida et al. [14] presumed that these high dosages were probably necessary to produce analgesia. But with our topical treatments, we were able to show that even low-dose peripheral applications can have a clearly analgesic effect. Furthermore, no skin changes were observed by us, even though corresponding case reports exist [31].

In view of this interesting therapeutic approach, a number of other Na_v 1.8-blocking substances are currently being developed. However, its highly potent blockade, good bioavailability and very few side effects make ambroxol a particularly interesting substance, and one that should prove at least as useful as other sodium channel blockers for the treatment of chronic pain [27]. The safety and tolerability of systemically administered ambroxol—even at dosages on the gram scale—have been demonstrated over decades of clinical experience.

Apart from case report 6, all patients suffered from hypersensibility to mechanical stimuli which regressed in each case after topical ambroxol. Is this anti-allodynic, analgesic effect really conceivable?

Allodynia and hyperalgesia are considered to be neuroplastic phenomena of the spinal sensory system [25]. The Na_v 1.8 channel is detected mainly in C- or A-delta fibres and neurons of the posterior horn [2, 6, 35, 52], although it is also expressed in A-beta fibres [1, 4, 26, 34, 40]. But since chronic inflammation shifts the excitability of Na_v 1.8 toward hyperpolarization this contributes to the allodynia, which means that ambroxol blockade would therefore be particularly pain-relieving. In animal studies, ambroxol also suppressed allodynia to differing degrees after systemic administration: heat hyperalgesia by 100 %, cold hyperalgesia and mechanical allodynia by approximately 75 % [14]. The intrathecal administration also showed an anti-allodynic effect in animal experiments [32]. The observation that ambroxol also managed to reduce mechanical allodynia in rats with experimental induced inflammation by approximately 2/3 [4] suggests that the anti-allodynic, analgesic effect is not necessarily limited just to neuropathic pain. In the patients described in this article, at least, the allodynia in response to mechanical stimuli was alleviated in all cases by topical ambroxol, which can now be explained by the rationale described above.

The onset of Na_v 1.8 blockade in vitro by ambroxol starts within a few seconds and is concentration-dependent and fully reversible [27]. In paraplegic rats hypersensitivity to static mechanical stimuli was reduced after approximately 30 min for approximately 3 h [16]. Both findings correlate closely with the statements of our patients (Fig. 2), according to which the effect started between 15 and 30 min after application. In case report 1, the effect was frequently even said to appear ‘immediately’ and persisted for well over 6 h in this patient. When applied repeatedly, topical lidocaine 5 % often shows an increasing effect over time, both clinically and according to the literature [29]. A comparable result was also reported by our patient 6, who felt that his underlying pain had almost disappeared after 4 months of use. The significance of these observations needs to be explored in greater depth.