9.1 Introduction

It is common in an orofacial pain center for a new patient to report how she or he has received substantial pain relief as a result of using one or multiple antibiotic medications. This story is not surprising if the patient was treated for a dental or sinus infection, but when you examine the patient, there is frequently no physical evidence of an infection. It is not uncommon for these patients to report that they have had unsuccessful root canal treatments on several teeth or even had several extractions in addition to multiple courses of various antibiotics. Moreover, when you inquire if the antibiotics provided relief for an extended period of time, the patient typically reports that as soon as they stop taking the antibiotics their pain returned to its original intensity. This clinical course has several possible explanations: (1) antibiotics reduce pain by suppressing local infection and associated swelling, which will indirectly reduce pain; (2) antibiotics might reduce pain directly via analgesic properties; or (3) antibiotics reduce pain via a placebo effect like all medications. However, there are adverse consequences in the chronic use of antibiotics to control pain.

9.2 Indirect Effect of Antibiotics on Tooth Pain Due to Reduction of Swelling and Inflammation

Pain reduction with antibiotics in the presence of infection is presumably due to decreased inflammation as the bacteria are killed. However, four prospective well-controlled trials in patients with painful pulpitis or periapical infections have failed to provide evidence of antibiotic efficacy for decreasing pain or decreasing analgesic consumption when compared with a placebo medication.^1–4 The results did not differ if the antibiotic (penicillin or amoxicillin) was given prophylactically (before) or immediately after the endodontic treatment was performed. Overall these studies suggest that prescribing antibiotics (penicillin or amoxicillin) for pain control is not effective for dental pulpal infections and the most important pain control method is removal of the infection itself. These findings do not explain the clinical observations that patients with chronic orodental pain report pain relief when they take antibiotics.

9.3 Known Antinociceptive Action of Various Antibiotics

One possible explanation for the analgesic effect of antibiotics is that these medications are not suppressing pain by reducing or killing bacteria, but they have direct analgesic properties. This issue has been examined by a series of experiments, mostly in vitro studies, described in the subsections focusing on the following antibiotic types: beta-lactams, aminoglycosides, tetracyclines, chemically modified tetracyclines (CMTs), and nucleoside antibiotics.

9.3.A Beta-Lactam Antibiotics

The initial demonstration of an innovative study examining the antinociceptive effects of antibiotics used an experi­mental rat model of pain created by sectioning dorsal roots C5 to T1 unilaterally.⁵ The rats demonstrated self-mutilation of the denervated limb that was quantified before and after the injection of four substances: a morphinomimetic drug and three antibiotics (chloramphenicol, amoxicillin, and doxycycline). The animals of the group treated with the morphinelike drug (pethidine) performed significantly less autotomy than did the animals in the control group. This same self-mutilation of the denervated limb was found when the animals were injected with chloramphenicol and amoxicillin, but doxycycline was found less efficacious. Another study in an acute pain model (hot plate) examining the effects of nine randomly selected antibiotics in rats⁶ also demonstrated that several antibiotics have antinociceptive properties. This antinociceptive effect was comparable to the effect produced with salicylate and ketoprofen. They reported that pain reduction was long lasting with chloramphenicol (10 hours or more) and the authors concluded that these antinociceptive properties cannot be attributed to sedation but are most likely due to analgesia.

9.3.B Aminoglycoside Antibiotics

Another series of studies in a variety of animal models have demonstrated antinociceptive effects of aminoglycoside antibiotics, including gentamicin, neomycin, kanamycin, and streptomycin.^7–18 The effects were dose-related and their relative antinociceptive potency was similar to the efficacy for blocking N-type calcium channels with anticonvulsant medications or neural blockade with local anesthetic. These studies provide ample preclinical evidence that aminoglycosides have real antinociceptive effects. Unfortunately, no published report has evaluated whether these antibiotics are analgesic in humans. There has been one report in the literature of inadvertent epidural infusion of gentamicin instead of fentanyl in a patient with back pain postpartum.¹⁹ The pain in this case was not controlled with this infusion, and in fact there was a return of pain that prompted the nurse to discover the inadvertent use of gentamicin.

9.3.C Tetracycline-Class Antibiotics (Minocycline and Doxycycline)

The effects of minocycline on activation of glial cells (microglia and astroglia) in the spinal cord of rats were examined after experimental nerve injury.²⁰ This study reported that intrathecal administration of minocycline, a selective inhibitor of microglial cell activation, inhibits low-threshold mechanical allodynia, as measured by the von Frey test, in two models of pain facilitation. In a rat model of neuropathic pain induced by sciatic nerve inflammation (sciatic inflammatory neuropathy), minocycline delayed the induction of allodynia in both acute and persistent paradigms. In addition, minocycline was able to stop the established inflammatory-induced allodynia at day 1, but not at 1 week later, suggesting a limited role of microglial activation in more-persistent pain states. This data is consistent with a critical role for microglial cells in initiating, rather than maintaining, enhanced pain responses. In addition to the above study, a 2006 study supports the concept that minocycline suppresses microglial-driven neuronal activity for at least 1 day after nerve injury.²¹ Additional studies demonstrate that minocycline suppresses serum levels of interleukin-6 (IL-6) in a chronic constriction injury (CCI) model of neuropathic pain in rats,²² suppresses microg­lial cells when given preemptively before an experimental nerve injury,²³ and prevents or delays the development of neuropathy.²⁴

In humans, there are three studies of importance to note. The first was a 1996 study that examined the anti-inflammatory effect of minocycline in rheumatoid arthritis (RA) subjects.²⁵ Serum samples of 65 RA patients who completed a 26-week randomized double-blind trial of minocycline (100 mg twice a day) versus placebo were studied. The authors reported that several clinical parameters and in particular the acute phase response markers of inflammation decreased significantly in the minocycline-treated group. Serum levels of IL-6, a proinflammatory cytokine decreased in the minocycline-treated group only and this decrease was positively correlated with the decrease in C-reactive protein (CRP) levels. Minocycline significantly decreased serum immunoglobulin-M rheumatoid factor (IgM-RF), IgA-RF, total IgM, and total IgA levels. No such changes were observed in the placebo-treated group. The anti-inflammatory effect of minocycline in RA patients may be due to the reduction in the synthesis of IL-6 and rheumatoid factor. Although this study did not measure pain outcome, minocycline could indeed play a role in pain relief by decreasing inflammation. Further studies with pain outcome will be useful to understand clinical impact of minocycline in pain relief in rheumatoid arthritis patients.

A second study in patients with osteoarthritis indicates that doxycycline slows the progression of osteoarthritis.²⁶ As determined by less joint space narrowing (JSN) in the drug group compared with the placebo group, pain scores in both treatment groups were low at baseline and remained low throughout the trial, thereby preventing determination of an analgesic effect in osteoarthritis patients. Finally, minocycline oral rinses were compared with a placebo aqueous solution mouthwash for assessing pain relief in patients suffering from frequent episodes of recurrent aphthous stomatitis (RAS).²⁷ Minocycline mouthwashes were reported to result in significant reduction in the severity and duration of pain (recorded by visual analog scale) due to RAS. Seven patients also participated in a crossover study and the findings were comparable with that of the randomized study. The authors concluded that minocycline oral rinses reduced pain in patients with RAS and may have implications for treatment in other noninfectious inflammatory ulcerative oral mucosal diseases. In summary, the minocycline and doxycycline studies just described show anti-inflammatory effects, which may reduce the pain levels. However, whether the pain reduction is actually due to the effects of these drugs needs to determined with further research.

9.3.D Chemically Modified Tetracyclines

Chemically modified tetracyclines (CMTs) do not have antimicrobial properties but maintain their function as matrix metalloprotein inhibitors (MMPs)^28,29 and are now being studied as anti-inflammatory agents³⁰ and antineoplastic agents.³¹ However, no analgesic efficacy testing of CMTs has been conducted in animals or humans. For example, Periostat^® is doxycycline hyclate modified to remove its antibacterial properties. Studies have shown that Periostat® inhibits or reduces collagenase activity in gingival crevicular fluids and periodontitis.^32–35 Periostat is given orally as a 20-mg tablet as an adjunct to scaling and root planing procedures to aid in periodontal tissue healing. However, no studies have been done to show analgesic activity of Periostat.

9.3.E Nucleoside Antibiotics

An incidental finding of the successful inhibition of neuropathic pain in a patient with long-standing Raynaud’s disease who received the trial chemotherapeutic agent KRN5500, a derivative of the nucleoside antibiotic spicamycin, led to further research which concluded that KRN5500 is effective against neuropathic pain but interestingly not against nociceptive pain.^36–38 Further research should shed more light on the mechanisms />