Dear Sir, As Drs. Greeley and Karst point out, the timing of subdural haemorrhage (SDH) is ‘an area which remains opaque and complex’. This is of concern, since oversimplified descriptions of the infant dura have led to clinical diagnoses and timing estimates that do not comport with the neuropathology. The only way to avoid these mistakes is through basic research, such as that presented in our paper. Through such research, we hope to bring clarity not only to the issue of timing of subdural collections, but also to their pathophysiological significance in the infant, which we have addressed separately [1, 2]. As noted, there is nothing new in the demonstration of the inflammatory response to SDH. Indeed, our finding that mast cell density increases with the age of the SDH is consistent with similar studies of eosinophils in adults [3]. We believe, however, that ours is the first study to examine this response in foetal and infant cases. We also agree that what we are observing is part of a generalised inflammatory response, and that the specific cell type involved may be less important than the fact that any inflammation—even an acellular ‘inflammatory soup’—is capable of both stimulating and sensitising dural trigeminal sensory afferent fibres, causing them to be activated by stimuli that previously evoked little or no response [4]. We drew particular attention to mast cells because they are rapid responders in the neurogenic modulation of immunity in host defence [5], and can be accurately identified through tinctorial and immunohistochemical stains, increasing reliability and facilitating counting. With regard to our methods, we noted that ‘[b]linding as to the nature of each slide (age and control versus subdural haemorrhage (SDH)) was maintained as far as possible throughout counting. ’ The assessors (AV, JD, and AG) had no clinical information and did not know whether they were counting control or test sections. While it was not possible to blind them to the presence of SDH, since this can be seen in the dura, the assessors were not neuropathologists and were not in a position to speculate on the age of the bleeding while conducting the counts. The statistical analysis was performed with advice from Professor Sarah Darby at the Clinical Trial Service Unit, University of Oxford. The data were analysed by constructing a variable to indicate the SDH status. The codes for this variable were 1 for controls, 2 for fresh SDH and 3 for chronic SDH. A linear regression analysis was then carried out relating mast cell density to SDH status using the computer program Stata version 12. The significance of the regression coefficient was evaluated using a two-sided t test. Drs. Greeley and Karst correctly note that it is not possible to reach statistical significance in a sample of this size. For this reason, we refrained from making such inferences, instead describing ‘borderline statistical significance’ and indicating that ‘[f]urther studies should more closely examine the relationship between age and mast cell density in the human dura, both in controls and in cases of subdural haemorrhage. ’ Because larger studies are more likely to be statistically reliable, we are not sure it makes sense to reformat our existing data, but if Drs. Greeley and Karst suggest any supplementary analyses we would be happy to run them. Drs. Greeley and Karst refer to an ‘apparent discrepancy’ in our observation that dural samples from older patients (adults, defined arbitrarily as over 200 months) tended to have a lower mast cell density than those from children, infants and A. Varatharaj (*) :W. Squier Department of Neuropathology, John Radcliffe Hospital, Oxford OX3 9DU, UK e-mail: a.varatharaj@gmail.com
Read full abstract
Nov 1, 2016
Can Yaldiz + 3
Open Access