To the Editor:
Sir, we read with interest the article by Walker et al. and the accompanying editorial view.1,2Undoubtedly, subarachnoid administration of large doses of ketamine produces neuronal apoptosis in newborn rats, as was eloquently demonstrated by this article. However, we would like to request further clarification regarding the statement “3 and 10 mg/kg produced increasing initial sedation, and higher doses were lethal.” Unlike the corresponding article regarding the safety of intrathecal morphine in rat pups in the same issue,3no indication of calculated LD50of intrathecal ketamine is given. We are not suggesting that similar dose response curves need to be constructed4,5but would welcome the publication of supporting data.
Rat pups were also exposed to smaller doses of intrathecal ketamine (0.1–0.3 mg/kg); again, no data on analgesic action or neuronal apoptosis are given. These doses (rather than more than 3 mg/kg) are the comparative and relevant equivalents commonly employed for caudal anesthesia.6
We have also some concerns regarding reporting of the apoptosis data.1First, the authors are assuming that the cells they are staining with active caspase-3 are indeed neurons without assessing the cell type. Second, the authors have failed to fully normalize the data in that they have just counted the number of positive cells, be they Fluoro-Jade C or caspase-3 positive, in each field of view. This method does not take into account the number of cells present in the section or the size of the section. One method is based on the generation of a “wandering mean.” To generate these data, the following procedure should be undertaken. Count the number of events (caspase-3 positive or Fluoro-Jade C positive cells) and the total number of relevant cells in the first microscopic field. This will give the first apoptosis score (A1 based on N1 cells). In the second field, the process is repeated and running scores recorded to give a running mean (A2 based on N2 cells). This process is repeated to give multiple running averages (A3, N3... An, Nn). If these are plotted, the mean will be seen to wander and eventually oscillate about a mean value, and as N increases, this will become less. This procedure can then define experimentally the number of events to be assessed to produce a given quality of data.7
Intrathecal ketamine may have a much narrower intrathecal therapeutic index compared with that of morphine. However, local anesthetic agents have been shown to have detrimental effects on neuronal apoptosis,8and the right balance between exposing vulnerable children to potential harmful general or regional anesthetics is yet to be established. Until then, we need to pay attention to the primary cause of morbidity and mortality in children: hypoxia.9
* Royal Aberdeen Children's Hospital, Aberdeen, United Kingdom. t.engelhardt@nhs.net
