Peritonsillar morphine infiltration to prevent early postoperative pain after tonsillectomy

Editor, A number of different surgical and medical techniques have been studied in the search for safe and effective post-tonsillectomy pain relief, but the results have been controversial.1 The peripheral administration of opioids has been found to potentially produce analgesia without the systemic side-effects in humans.2,3 The aim of the present study was to evaluate the effectiveness of peritonsillar infiltration of morphine in attenuating early postoperative pain after tonsillectomy in adult patients. A randomised, controlled trial was carried out after obtaining the approval from the ethics committee of the Northern Ostrobothnia Hospital District (Ethical Committee N° 106/2004) on 1 October 2004. Based on a previous study with a similar methodology,4 we found that 30 patients would be required to detect a difference of 10% in pain scores with a 90% power and 0.05 significance level. The patients served as their own controls. Written informed consent was obtained from 40 adult patients undergoing elective bilateral tonsillectomy because of chronic or recurrent tonsillitis between 16 March 2005 and 11 June 2007 in the Department of Otorhinolaryngology, Oulu University Hospital, Oulu, Finland. Patients with history of peritonsillar abscess and sleep apnoea syndrome, and those with present acute infection or major organ disease were excluded. Blunt dissection tonsillectomy was performed under standardised general anaesthesia. The patient's side for peritonsillar infiltration was randomly allocated using sealed instructions from an envelope. At the end of the operation, peritonsillar tissue was infiltrated with 4 mg (4.5 ml) morphine on one side (morphine side) and 4.5 ml 0.9% NaCl (NaCl side) on the other side in the superior, middle and inferior poles of the fossa. Neither the patient nor the surgeon and postoperative observer knew the side on which peripheral morphine had been injected. Ketoprofen (100 mg) was intravenously administered at the end of the operation as a routine. Fentanyl 25 to 50 μg was given in intravenous boluses as needed in the recovery room. Ketoprofen and paracetamol along with codeine combination were used as basic per oral analgesics on the ward and at home. A numerical rating scale (NRS: 0 no pain, 10 worst pain imaginable) was used to evaluate postoperative pain. Pain at rest (in the absence of any pharyngeal movement) and on swallowing were recorded for each operation side at 15 minutes interval during the first 2 hours, every hour to next 4 hours, and twice a day 2 weeks after the operation. Primary outcome was the severity of postoperative pain as assessed by NRS scores at rest and on swallowing. NRS pain scores over time are displayed as median (interquartile range). A linear mixed model for repeatedly measured data was applied for statistical comparison in NRS between morphine and NaCl sides over time. The P values reported with linear mixed model are Ptime for the change over time, Pgroup for the average difference between sides and Ptime × group for the interaction between time and side. Forty patients with a mean age of 30 years (SD 7.3) and mean weight of 74 kg (SD 9.6) were included in the study. However, 29 patients (10 men) returned the 2-week follow-up diary, and thus, 11 patients (seven men) could not be analysed for the entire follow-up period. Morphine was injected on the left side in 16 cases and on the right in 13 cases. Pain scores at rest and on swallowing decreased during the first 6 h after surgery, but there was no evidence for a difference between the morphine and NaCl sides (Fig. 1a and b). In addition, there was no evidence for a difference between the sides in the NRS pain scores at rest and on swallowing throughout the study period. All patients received intravenous fentanyl during their stay in the recovery room or on the ward. The mean fentanyl consumption was 57 (SD 44) μg in all patients. No serious or unexpected adverse events occurred.Fig. 1: Median (interquartile range) numerical rating scale (NRS) pain scores (a) at rest and (b) on swallowing during the operation day (O – 6 h) on the morphine side (▪) and on the NaCl side (□). No statistically significant difference between the sides. The values are presented as medians with 25th and 75th percentiles.We found that the local infiltration of 4 mg morphine at the end of the tonsillectomy had no effect on the severity of pain during the early and long-term postoperative follow-up on the injected side. A dose of 4 mg morphine was chosen to avoid any systemic effect, while ensuring a good local spread of the molecule with effective peripheral effect. The presence of similar pain scores on both sides with morphine and NaCl may be explained by three different hypotheses. First, the peripheral infiltration of a 4 mg morphine dose may had exerted an effect peripherally on both the sides simultaneously and potentially decreased peripheral and central sensitisation. We cannot exclude, however, a potential central analgesic effect resulting from systemic absorption because of the intense vascularity of the peritonsillar tissue. Second, perioperative and immediate postoperative use of fentanyl may have disguised the analgesic effect of the peripheral morphine. Khalefa et al.5 found that in the presence of inflammatory pain, the antinociception of systemically administered fentanyl and morphine greatly surpasses the antinociception effect obtained after local application through the activation of supraspinal and to a lesser degree spinal, but not peripheral μ-opioid receptors. Therefore, the present study may also suggest that peripheral opioid analgesia may be limited in its maximal effect as compared with the systemically absorbed morphine and fentanyl which were used in the perioperative and early postoperative period. Finally, Godai et al.6 suggest that local administration of morphine attenuates the development of post-incisional pain through macrophage-dependent mechanisms at the incisional sites. The morphine-induced analgesic effects, however, were reversed by local administration of a nonselective cyclo-oxygenase (COX) inhibitor, indomethacin. Owing to the fact that morphine increased the expression of prostaglandin E2 (PGE2), and the analgesic effects of μ-opioid receptor were inhibited by a COX-2 inhibitor, indomethacin, the analgesic effects of μ-opioid receptor signalling might be dependent on COX-2/PGE2. Therefore, in our study, infusion of the nonselective COX inhibitor, ketoprofen, might have alleviated the peripheral morphine effect. In conclusion, this study did not provide evidence of the benefit of peripheral injection of an opioid, indicating that it might not be an optional tool to prevent pain in daycare tonsillectomy. Although this study may have some limitation because of the relatively modest dose of morphine, higher concentrations may have caused non-negligible systemic effect and brought into question the benefits of the peripheral opioid effect. The evidence for using the peripheral or systemic morphine injection after tonsillectomy is still unclear and needs to be substantiated in further prospective controlled studies. Acknowledgements relating to this article Assistance with the study: none. Financial support and sponsorship: none. Conflicts of interest: none.