Adenosine analogs injected intrathecally (IT) in rodents induce antinociceptive effects in a variety of nociceptive tests [1-4]. Together with findings of biochemical markers for adenosine formation, release and metabolism in the dorsal horn, and the existence of spinal adenosine receptors, this indicates a role for adenosine in the modulation of nociception at the spinal level . Of special interest are reports on the effect of adenosine analogs injected IT in tests mimicking neurogenic pain states [2-4]. Using IT strychnine, Sosnowski and Yaksh  have demonstrated that adenosine analogs inhibit the pain behavior induced by the glycine antagonist strychnine. Adenosine analogs reduce the pain behavior after IT prostaglandin F2 alpha, which induces a reaction similar to IT glycine . The hyperalgesia induced by incomplete sciatic nerve lesion is antagonized by IT adenosine analogs . Sklar et al.  showed that adenosine monophosphate injected intramuscularly in patients with postherpetic pain was effective in controlling the postinfectious neuralgia. These studies indicate a role for adenosine analogs in the treatment of neurogenic pain. In clinical treatment of neurogenic pain, current drug therapy is usually not successful in severe cases. There is a clear need to develop new pharmacologic tools for the treatment of such patients. The spinal route of administration is of primary interest, since the antinociceptive effects of adenosine analogs are probably mediated by spinal mechanisms  and adenosine, administered systemically, has been shown to induce pain in animals [7,8] and humans [9,10].
Phenylisopropyl adenosine has been administered to healthy human volunteers in a study on the antilipolytic effects of adenosine analogs . From 5 to 15 micro gram/kg was given orally or intravenously, and the authors reported mild side effects consisting of short-lasting oppression in the gastric region and heaviness and numbness of the limbs. To our knowledge, adenosine analogs previously have not been administered IT to humans to treat pain. Extracellular adenosine receptors are divided into A1 and A2 subtypes and several analogs have been developed with different affinity for these receptors. In a series of experiments we have shown that A1 adenosine agonists are possible candidates as future analgesic drugs [1,8,12]. In safety screening tests, IT injection of the A1 agonist R-phenylisopropyl adenosine (R-PIA) induces a slight increase in spinal cord blood flow . After chronic administration of R-PIA, no histologic changes could be detected in the spinal cord in rats as detected by light and electron microscopy combined with morphometric analysis .
In this case report we present the results of IT administration of R-PIA in one patient with a well characterized intractable neurogenic pain state. The treatment was approved by the ethics committee of the Medical Faculty of Uppsala University and was performed in conformity with the Helsinki Declaration after consent by the patient. R-PIA was solved and diluted in 0.9% NaCl to a concentration of 50 nmol/mL. The drug was prepared by our hospital pharmacy. It was sterilized by heat and the stability of the compound was confirmed by nuclear magnetic resonance spectral analysis. The prepared drug was controlled for sterility and for the presence of pyrogenic substances. The dose given to the patient was based on previous animal experiments  and extrapolated by weight.
The patient was a 47-yr-old man with a more than 10-yr history of severe neurogenic pain and allodynia to touch because of a nerve lesion after surgery for an ankle joint injury. The pain was localized to the distal lateral part of the lower right leg and extending to the dorsolateral surface of the foot (sural nerve). He was unsuccessfully treated with transcutaneous electrical nerve stimulation, guanethidine blocks, carbamazepine, antidepressant drugs, and analgesics including oral opioids. The patient also had a nociceptive pain in the foot when standing which was relieved by oral analgesics. However, the constant neurogenic component with allodynia to touch and vibration was severe and distressing for the patient.
Twenty-four hours prior to the injection of R-PIA the patient was instructed not to drink any tea or coffee, since xanthines antagonize the effects of adenosine analogs. His only medication was paracetamol and dextropropoxyphene, and this was omitted on the morning of treatment with R-PIA. Before the injection of R-PIA and every half-hour for up to 3 h after the injection, the patient was tested and asked to rate his pain score (visual analog scale [VAS] expressed as percent of maximal pain) both at rest and after light touch in the area of pain. Thermal sensitivity thresholds (Thermotest Registered Trademark; Somedic Sales AB, Farsta, Sweden) were assessed at the dorsolateral surface of both feet, using the uninjured side as control. Each value represents a mean of five measurements. Vibration thresholds (Vibrameter Registered Trademark; Somedic Sales AB), expressed in terms of amplitude of the stimulator movement (micro meter), were also measured at the dorsolateral surface of each foot. The mean of three measurements was recorded. The skin temperature was recorded using an infrared temperature scanner (Dermatemp Trademark; Exergen Corporation) at five points at the dorsal surface in each foot and the means of these measurements were calculated. The tendon reflexes in the legs were evaluated using a scale from absent (-) to strong (+++), and the ability to walk was noted. Electrocardiography was monitored continuously as was oxygen saturation. Respiratory rate and arterial blood pressure were measured every 30 min. His physical status is shown in Table 1. All measurements were repeated 24 h after the injection of R-PIA. Prior to the injection of R-PIA, 1 mL of NaCl 0.9% (single blinded) was injected IT. This had no effect on the patient's pain rating or nociceptive thresholds (data not shown).
Using a 27-gauge spinal needle, R-PIA (50 nmol) was injected at the L3-4 level with the patient in the sitting position. He did not experience any discomfort during the injection. IT injection of R-PIA decreased the allodynia elicited by vibration Figure 1 and touch within 20 min and also reduced the resting pain Figure 2 as assessed by VAS. The threshold for detection of warmth and cold was unchanged in the feet, as were the thresholds for pain sensation elicited by warm stimuli Figure 3. Cold stimuli did not induce pain either before or after the injection of R-PIA. No motor disturbances occurred; the patient could walk normally, and tendon reflexes were normal Table 1. Pulse rate, blood pressure, oxygen saturation, and respiratory rate were unchanged Table 1. Three hours after the injection of R-PIA, the patient reported a feeling of warmth in his feet lasting approximately 24 h. A slight increase in skin temperature was noted in the feet Figure 4. Twenty-four hours after the IT injection of R-PIA the allodynia to touch and vibration was absent (VAS 0) and the perceptive thresholds for warmth and cold were still unchanged. After 10 days the constant resting pain began to recur, and 2 wk after the injection of R-PIA it returned to preinjection level. However, the allodynia to touch and vibration did not return (6 mo observation). The pain that the patient experienced while standing (nociceptive component) was not affected by IT R-PIA.
This case report substantiates the hypothesis that adenosine is involved in the modulation of pain. The most probable site of action, as supported by animal experiments, is the dorsal horn. The patient, with a well defined neurogenic pain component, including allodynia for touch and vibration, had a decreased pain rating, and the allodynia was eliminated completely after IT R-PIA. This implies that adenosine analogs might affect the mechanisms behind allodynic pain states. From this case we can draw conclusions concerning only low threshold mechanical stimulation, since the patient had no allodynia evoked by cold or warm stimuli. No effects were seen on the nociceptive response to thermal stimulation at the dose tested (50 nmol), indicating that normal nociception and sensory processing was not affected by IT R-PIA.
The effect on the allodynia was surprisingly long-lasting. We can only speculate on the reasons for this unexpected finding. It may imply an effect of R-PIA on pathologic neuronal activity in the dorsal horn of the spinal cord. A single dose of R-PIA may have affected the neuronal plasticity and thereby induced a functional modulation of neuronal activity, extinguishing the hyperexcitability. A pure pharmacologic effect on specific receptors would have been of shorter duration. The knowledge of the spinal pharmacokinetics of R-PIA is limited, and detailed studies are needed before any conclusions about the long-lasting analgesic effect can be drawn.
According to previous animal studies, side effects can be anticipated. Central respiratory depression has been observed after intracerebroventricular injection of phenyisopropyl adenosine in anesthetized rats . Adenosine analogs administered intravenously to conscious rats have a depressant effect on the systemic circulation . No such side effects were noted in our patient. His respiratory frequency and oxygen saturation was not affected, nor was his pulse rate or blood pressure. Of course, this might mean only that the present dose (50 nmol) was too small to induce systemic or supraspinal effects and cannot be taken as evidence that IT R-PIA in larger doses is without harmful effects. R-PIA is considered to be lipophilic  and this might reduce the possibility for supraspinal effects after IT administration. During the followup of the patient no adverse effects were seen.
We are well aware of the scientific difficulties and limitations of a single case study. However, some interesting indications on the general effects of IT R-PIA can be noted. IT R-PIA had no effect on thermal nociception but abolished the allodynia evoked by touch and vibration in a patient with neurogenic pain. Of course, further clinical studies are needed to define the potential role for IT adenosine analogs in chronic pain treatment. This should be done with care, since the toxicologic data are sparse. It is also important that the patient's pain states be well defined.
1. Karlsten R, Gordh T Jr, Hartvig P, Post C. Effects of intrathecal injection of the adenosine receptor agonists R-phenylisopropyl-adenosine and N-ethylcarboxamide-adenosine on nociception and motor function in the rat. Anesth Analg 1990;71:60-4.
2. Sosnowski M, Yaksh TL. Role of spinal adenosine receptors in modulating the hyperesthesia produced by spinal glycine receptor antagonism. Anesth Analg 1989;69:587-92.
3. Minami T, Rumiko U, Horiguchi S, et al. Allodynia evoked by intrathecal administration of prostaglandin F2
alpha to conscious mice. Pain 1992;50:223-9.
4. Yamamoto T, Yaksh TL. Spinal pharmacology of thermal hyperesthesia induced by incomplete ligation of sciatic nerve. I. Opioid and nonopioid receptors. Anesthesiology 1991;75:817-26.
5. Sawynok J, Sweeney MI. The role of purines in nociception. Neuroscience 1989;32:557-69.
6. Sklar SH, Blue WT, Alexander EJ, Bodian CA. Herpes zoster. The treatment and prevention of neuralgia with adenosine monophosphate. JAMA 1985;253:1427-30.
7. Taiwo YO, Levine JD. Direct cutaneous hyperalgesia induced by adenosine. Neuroscience 1990;38:757-62.
8. Karlsten R, Gordh T, Post C. Local antinociceptive and hyperalgesic effects in the formalin test after peripheral administration of adenosine analogues in mice. Pharmacol Toxicol 1992;70:434-8.
9. Sylven C, Beermann B, Jonzon B, Brant R. Angina pectoris-like pain provoked by intravenous adenosine in healthy volunteers. Br Med J 1986;293:227-30.
10. Sylven C, Jonzon B, Fredholm BB, Kaijser L. Adenosine injection into the brachial artery produces ischaemia like pain or discomfort in the forearm. Cardiovasc Res 1988;22:674-8.
11. May B, Leinweber W. Antilipolytic action of phenylisopropyl adenosine in man. Naunyn Schmiedeberg's Arch Pharmacol 1971;269:466-7.
12. Karlsten R, Post C, Hide I, Daly JW. The antinociceptive effect of intrathecally administered adenosine analogs in mice correlates with the affinity for the A1-adenosine receptor. Neurosci Lett 1991;121:267-70.
13. Karlsten R, Kristensen JD, Gordh T Jr. R-phenylisopropyl adenosine (R-PIA) increases spinal cord blood flow after intrathecal injection in the rat. Anesth Analg 1992;75:972-6.
14. Karlsten R, Gordh T Jr, Svensson BA. A neurotoxicologic evaluation of the spinal cord after chronic intrathecal injection of R-phenylisopropyl adenosine (R-PIA) in the rat. Anesth Analg 1993;77:731-6.
15. Hedner T, Hedner J, Wessberg P, Jonason J. Regulation of breathing in the rat: indications for a role of central adenosine mechanisms. Neurosci Lett 1982;33:147-51.
16. Milavec-Krizman M, Wagner H, Kralj A, et al. Cardiovascular effects of adenosine derivatives in conscious normotensive rats. Nucleosides Nucleotides 1991;10:1127-9.
© 1995 International Anesthesia Research Society
17. Vapaatalo H, Onken D, Neuvonen PJ, Westermann E. Stereospecificity in some central and circulatory effects of phenylisopropyl-adenosine (PIA). Arzneimittelforschung 1975;25:407-10.