Effects of gabapentin on morphine consumption and pain in severely burned patients

1. Introduction 

Burn patients often suffer unbearable pain that is difficult to control. Although nociception and peripheral hyperalgesia are considered the major causes of burn pain, mechanisms like central hyperalgesia and neuropathic pain are important components of acute burn pain [1]. Animal and volunteer studies confirm a central hyperalgesia after burn [2]. Recent findings support the hypothesis that this post-burn hyperalgesia shares a common mechanism with the post-operative pain hypersensitivity [3]. Central neural sensitization probably contributes to some aspects of both type of pain; thus, Woolf suggested using antihyperalgesic drugs to improve treatment [4]. Gabapentin (Gp) is an antihyperalgesic drug that selectively affects the nociceptive process involving central sensitization [5]. Multicenter controlled trials have demonstrated analgesic efficacy and safety of Gp in chronic neuropathic pain [6]. The use of Gp in the perioperative setting has been evaluated in recent studies and demonstrated a reduction in morphine consumption and movement-related pain [7], [8]. Although a volunteer study showed that Gp reduces hyperalgesia following experimental burns [9], no study has yet concentrated on severely burned patients.

2. Materials and methods 

3. Results 

Two patients from the Gp group failed to complete the study. The 10 remaining patients were able to complete the study, and 10 matching patients were retrospectively chosen out of a period of 42 months.

The demographic data TBSA, and burn locality were similar among the two groups (Table 1).

Table 1. Patients characteristics

		Gabapentin	Control
	Age (years)	45±14	46±14
	Sex ratio (M; F)	10; 0	7; 3
	TBSA (%)	25 [12–50]	28 [15–59]
	Weight (kg)	83±16	78±18
	Burns localization (number)
	Head	10*	2
	Back	10	10
	Buttocks/genetelia	3	4
	Upper limbs	9	6
	Hands	3*	6
	Lower limbs	8	7

Results are mean±S.D. or [range]; TBSA, total burn surface area (n=10 in each group).

* p<0.05, Chi-square test.

The number of operations was similar among the two groups (Table 2).

Table 2. Number of total operative cases and details over procedures

		Treatment phase		Post-treatment phase
		Gabapentin	Control	Gabapentin	Control
	Total	26	25	22	33
	Homografts	15	20	7	13
	Integra	3	0	0	0
	Autografts	14	10	17	23

Treatment phase, from 3rd to 23rd day post-injury; post-treatment phase, from 24th to 45th day post-injury (n=10 in each group).

Morphine requirements were significantly smaller in the Gp group during both the treatment and the post-treatment phases with Gp (Table 3).

Table 3. Cumulative morphine requirements in mg (mean±S.D.)

		Treatment phase		Post-treatment phase
		Gabapentin	Control	Gabapentin	Control
	Day 7	537±278	692±243	345±147	632±324
	Day 14	998±296	1286±408	636±270	1232±686
	Day 21	1325±372	1822±538	887±337	2017±1048

Treatment phase, from 3rd to 23rd day post-injury; post-treatment phase, from 24th to 45th day post-injury (n=10 in each group); p<0.05, for each reported data.

The daily pain scores were similar among the groups during the first 2 days of treatment (Fig. 1a). During the rest of the treatment phase however, pain scores became significantly smaller in the Gp group: VAS between a minimum of 1.5 [2] and a maximum of 4 [1] for the Gp group versus VAS between a minimum of 4 [2] and a maximum of 7.5 [6] for the control group. The daily pain scores post-treatment were significantly lower in the Gp group at days 1 and 2, 4 and 5, and from days 8 to 17 (Fig. 1b). During the whole post-treatment phase, pain scores in the Gp group kept being low and varied less than those in the control group: VAS between 2 [2] and 3 [4] for the Gp group versus VAS between 3 [3.5] and 7 [4] for the control group.

View Large Image Download to PowerPoint

Fig. 1. Mean daily pain scores. (a) During the 21 days of treatment with 800mg of oral gabapentin three times a day (2400mg/day). (b) During the 21 days after stopping the treatment with gabapentin. VAS 0–10: 0, no pain; 10, worst pain possible. Gabapentin: gabapentin group (n=10). Control: retrospective group (n=10). The boxes represent the 25th–75th percentiles, the dark line is the median; the extended bars represent the 10th–90th percentiles, and the circles represent values outside this range.

A total of seven occurrences of side effects related to Gp were noted in subjects during the treatment phase (Table 4). The commonest side effects were somnolence and dizziness. Somnolence was present in three patients during Gp treatment but did not preclude them from continuing to take this medication. All side effects resolved during the post-treatment phase except for two patients reporting somnolence during the 1st week post-treatment. We could not compare these side effects of Gp with the control population as these were specific to Gp. A comparison of opioid-induced nausea and vomiting (Table 4) shows a higher incidence of nausea and vomiting during the treatment phase in the control than in the Gp group (four patients versus one patient, p<0.05). Constipation and itching were common in both groups during both phases of the study.

Table 4. Side effects

		Treatment phase		Post-treatment phase
		Gabapentin	Control	Gabapentin	Control
	Gabapentin-related
	Somnolence	3	ND	2	ND
	Lightheadedness	1	ND	0	ND
	Dizziness	2	ND	0	ND
	Headache	1	ND	0	ND
	Visual disturbances	0	ND	0	ND
	Ataxia	0	ND	0	ND
	Morphine-related
	Nausea/vomitting	1*	4	0	0
	Pruritus	6	5	10	10
	Constipation	2	3	4	4
	Urinary retention	0	0	0	0
	Respiratory depression	0	0	0	0

Results are numbers (n=10 in each group); ND, not determined; treatment phase, from 3rd to 23rd day post-injury; post-treatment phase, from 24th to 45th day post-injury.

* p<0.05, Chi-square test.

4. Discussion 

We administered 2400mg to our patients in three divided doses. This regimen produces median plasma levels of Gp around 40μmolml−1 [10], similar to those observed in healthy volunteers receiving Gp for burn pain [9]. A preliminary study – not published – taught us that patients tolerated 2400mg of Gp well. Gp has a favorable pharmacokinetic profile for burn patients [11]: it does not bind with plasma proteins; it has no significant drug interactions and is not metabolized in humans. It may be given by mouth with an absorption kinetics dose-dependent, due to a saturable transport system. Elimination is by renal clearance.

This study revealed a significant opioid-sparing Gp, administered for 21 days in severely burned patients, with minor adverse effects. This opioid-sparing effect persisted for the next 21 days, exceeding the pharmacologic duration of Gp.

Evidence for a central mechanism in burn-induced pain are numerous. The persistent activation of small C-fibers afferents with burn initiates a state in which the spinal neuron shows enhanced responses to low- and high-intensity stimuli [12]. Recently, a major role played by the spinal AMPA-KA receptors in the initiation of burn-induced secondary tactile allodynia was reported in rats [13]. In healthy volunteer studies, systemic antagonists of NMDA-receptors, acting on the dorsal horn of the spinal cord, reduced post-burn hyperalgesia [14], [15]. Gp is an antihyperalgesic drug that selectively affects the nociceptive process involving a central sensitization [16], [17]. Evidence suggests that the antihyperalgesic action of Gp in burn pain is likely to occur at the spinal level [18], [19]. Pharmacologic effects of Gp that may be important in pain suppression include binding to α2δ calcium channel subunits, suppression of glutamate and substance P transmission, as well as γ-aminobutyric acid receptors modulation. However, the relative importance of these mechanisms remains uncertain [17].

The diminished mean daily pain scores with Gp reported here may be explained by a reduction of a central neuronal hyperexcitability induced by painful procedures. The mean daily pain scores diminished only after the 3rd day of Gp administration and subsequently remained stable (Fig. 1a). An explanation of this delayed effect may be that Gp acts only on the component of pain produced in the spinal cord that leads to hyperalgesia, so an effect would only be apparent once this condition has developed [20]. Pain scores of the control group, on the contrary, were higher and far more variable.

Opioids are the mainstay of analgesia in burn pain. However, while effective for the background pain, they are reported to perform poorly for procedural pain [21]. Not surprisingly, morphine consumption was here mainly reduced among the patients receiving Gp, through a reduction of rescue analgesia after procedures during the whole the study. The reduction of pain and of morphine consumption means not only that the patients were more comfortable and awake but also that they could take a more active part in their rehabilitation. We deliberately chose to report the total cumulated morphine consumption rather than the daily morphine consumption for the sake of clarity. We felt that in a small number of patients, daily morphine consumption was not practical as surgery or dressing changes happened in the group on different days. Also, cumulative morphine doses can be interesting because high doses of opioids have immunological and endocrine effects: an inhibition of the cellular and humoral immunologic responses [22], [23], a hypocortisolism, and a growth hormone deficiency [24].

Also, increased doses of opioids cause more side effects. In our series, the increased incidence of nausea or vomiting reported in the control group (four in the control group versus one in the Gp group) is related to larger doses of morphine (Table 4). Pruritus was reported by every patient, but it is a constant symptom during the healing of burn wounds.

What we did not expect was an opioid-sparing effect exceeding the pharmacologic duration of Gp. Limitation of morphine tolerance may be a first mechanism [25]. A reversal of morphine tolerance by Gp may be an explanation, through an effect on the dorsal horn neuronal response [26], [27]. Opioid exposure activates descending pain facilitation mechanisms, enhancing nociceptive transmission at the spinal level and promoting hyperalgesia [28].

Multicenter studies about Gp in post-herpetic neuralgia [29] or diabetic neuralgia [30] reported somnolence (27%) and dizziness (24%) as the most frequent side effects. The most frequent Gp-related side effect among our patients was somnolence but all Gp-related side effects resolved before the end of the treatment phase.

The length of stay of the studied population was important. These patients all suffered from full-thickness burns of the occipital area that were slow to heal—and prevented any discharge prior to day 45 of hospitalisation.

In conclusion, a daily oral dose of 2400mg Gp for 21 days improved the analgesia and reduced opioid requirements in burn patients for a period effect exceeding the pharmacologic duration of Gp. However, the small number of patients in this pilot-study warrants randomized multicenter studies to confirm our observations.