In Reply

We would like to thank Drs. Youngblood and Harbott for their interest in our article¹  and for their valuable comments. In their letter, they voiced concerns about the methodology and interpretation of our results. At first, we have to mention the issue on the discrepancy in postoperative fentanyl consumption between current study and our historic control²  because Drs. Youngblood and Harbott believe that inconsistency in opioid consumption is one of the important evidences of withdrawal from remifentanil. As we had briefly described in the method sections of both studies, the surgical procedures were different from each other; the one in the former study was open ureteroneocystostomy via a 4–5 cm of Pfannenstiel incision (open Cohen technique), whereas the one in the current study was laparoscopic Cohen surgery under pneumovesicum. Although current laparoscopic technique has not yet achieved widespread acceptance in comparison with open procedure, minimally invasive surgical procedures have been increasingly replacing their open surgery counterparts in the field of pediatric urology.³  As laparoscopic ureteroneocystostomy usually requires only three of 3–5 mm of trocar entries, it has several advantages, including improved cosmesis and relatively reduced bladder trauma. Postoperative pain, of course, will be expected to be less than that of open procedure. Based on our empirical observations, the setting of patient-controlled analgesia in the former study has higher rates of background infusion by 25% than that in the current study (0.25 µg·kg⁻¹·h⁻¹ fentanyl vs. 0.20 µg·kg⁻¹·h⁻¹ fentanyl). Furthermore, if the children seemed to be consistently uncomfortable with initial settings, the fentanyl dose in the former study was doubled (56.3% of children were given “double fentanyl” during the first postoperative 48 h), whereas which was not needed in the current study. In addition, the age populations of both studies are considerably different as well; 0.5–2 yr versus 1–5 yr. As there has been no established pharmacokinetic/pharmacodynamic model of fentanyl especially in pediatric population, it is plausible to say that efficacies of µ-opioid agonists in children can be inferred by extrapolating from generally accepted remifentanil model. Minto et al.⁴  identified that age is a significant covariate of EC₅₀ in remifentanil model at the rate of 1.0–8.0 µg·kg⁻¹·min⁻¹ (EC₅₀ = 19.0 − 0.148 × Age). With increasing age, EC₅₀ decreased, suggesting that as age decreases, the potency of µ-opioid agonist decreases. Therefore, it seems unreasonable to assert remifentanil withdrawal based on the discrepancies seen in fentanyl consumptions between two apparently different studies.

Another important issue that Drs. Youngblood and Harbott had brought out is bridging dose of analgesics. We agree with their opinion in that absence of long-acting opioids may place those children at risk of postoperative pain. However, we wish to point out that the most prominent feature of pain after ureteroneocystostomy is moderate-to-severe intermittent pain due to not only surgical incision but also bladder spasms. Bladder spasm is a common adverse event after surgeries that involve bladder dissection. As the nature of pain is mainly intermittent, the patient-controlled analgesia response to postoperative pain is both quick and efficient. We regarded residual depressant effects of long-acting opioids as particularly harmful to the convalescent children after laparoscopic ureteroneocystostomy because the risk of laryngeal edema is higher than that of open procedure, owing to surgical positioning and prolonged operative time. Therefore, all children received patient-controlled analgesia for postoperative pain control rather than a bolus injection of long-acting opioids as the latter when administered just before or after the end of surgery may induce severe, delayed respiratory depression.⁵  Moreover, as children undergoing urologic surgery are not infrequently combined with various concealed anomalies, we respectively do not believe that routine use of long-acting opioid at emergence is always warranted, even in the absence of current trial. However, basal infusion of fentanyl for patient-controlled analgesia might attenuate the differences in the postoperative cumulative fentanyl consumptions in our study, although we of course considered that basal infusion would make children more comfortable.

According to current guidelines of opioid dosage,^† the recommended dose of remifentanil used to maintain sevoflurane anesthesia in children aged 1–12 yr is 0.05–1.3 µg·kg⁻¹·min⁻¹. In adults or adolescents, a remifentanil infusion dose of 0.3 µg·kg⁻¹·min⁻¹ is associated with the development of acute tolerance. In small children, precise dosage and duration of remifentanil that is associated with tolerance had not been determined, and data were fragmentary and conflicting. Despite our current result, we cannot conclude that administration of remifentanil at a dose of 0.3 µg·kg⁻¹·min⁻¹ does not cause acute tolerance after any duration of infusion. As tolerance to opioid is dose-dependent,⁶  part of the explanation for no hemodynamic evidence of waning remifentanil effect seen in our result may be related with age-related pharmacodynamic difference of µ-opioid agonist and possibly with discrepancy of infusion duration. We do not disagree with Drs. Youngblood and Harbott that our study was underpowered to assess opioid withdrawal thoroughly, and acute tolerance may not entirely fit with our results; every recovery profiles specific to opioid withdrawals were not evaluated in the immediate postoperative period, and the assessment of fentanyl consumption at earlier time points were not derived systemically as well. In the paucity of pediatric data on this topic, however, all the parameters relevant to opioid tolerance and withdrawal cannot be considered to be investigated before the study design. Furthermore, vast number of primary comparisons and the degree of multiplicity will create a large interpretation problem with regard to type-I error control (six repeated measurements for each parameter of four groups). Our report would be of greater value if we were able to show more detailed agitation score and continued sedation in immediate postoperative period. Perhaps we can attempt to do so in our next study as this issue is beyond the primary goal of the current study.