B. Blondel

Embolie graisseuse au cours des allongements de membres par clou centromédullaire : à propos de trois cas

Several methods are available for progressive limb lengthening, including centromedullary nailing, external fixation, or a combination. Each technique has its own advantages and drawbacks. In trauma victims, use of centromedullary nailing is associated with potentially fatal fat embolism. This fatal outcome might also occur during limb lengthening, particularly in bilateral procedures. To our knowledge, fat embolism has not been reported with the use of centromedullary nail for limb lengthening. This was a multicentric study of three cases of fat embolism, including one fatal outcome. In all, 36 centromedullary lengthening nails were inserted in the three centers before these acute episodes. The first two cases occurred during single-phase bilateral procedures, the third during unilateral lengthening. Fat embolism could result from several factors, as reported in the literature. While the bilateral nature of the procedure has been incriminated, the observation of an embolism during a unilateral procedure suggests other factors may be involved. Considerable increase in endomedullary pressure during reaming and insertion of the nail has been demonstrated. At the same time, there is the question as to whether the reduction of the diminution of medullary pressure by corticotomy would be an efficient way of reducing the risk of fat embolism. Based on the analysis of our three cases, we suggest that the best way to avoid fat embolism might be to drill several holes within the area of the osteotomy before reaming, in order to reduce endomedullary pressure. This can be achieved via a short skin incision, sparing the periosteum before low energy osteotomy. Since applying this protocol, the three centers have implanted 17 lengthening nails, without a single case of fat embolism.

Response to the letter by Vedran Deletis, David B. Mac Donald, Francesco Sala and Isabel Fernandez Conejero.

We wish to thank the authors of this Letter to the editor for their nterest in our article. We are pleased that members of the Interational Society for Intraoperative Neurophysiology (ISIN) support uch of the content and wholeheartedly embrace our concern egarding the value of remote monitoring and automated IOM evices. However, since some objections have been raised about ata and statements in this article [1], we wish to present our rguments point by point. 1. We agree with ISIN members that IOM must be performed as ontinuously as possible, in order to assess the functional integrity f neural structures at risk during critical surgical steps. However, rom a practical point of view, IOM, whatever the technique used SSEP, MEP, NMEP or D-waves) cannot be performed on a strictly ontinuous basis, notably during the use of diathermy knife. We ecall that some neurologic complications may have occurred while o instrumented correction was being performed. We have thus ndicated that IOM consists of subcontinuous evaluation of spinal ord sensorimotor functions during surgery, meaning of course as ontinuously as possible. 2. NMEP (neurogenic mixed evoked potentials) are used not only n France but also by other teams in charge of severe paediatric pinal deformities, for example in USA [2,3]. Finally, there is only ne clinical article that describes NMEP false negatives [4]. Conersely, two articles describe MEP (motor evoked potentials) false egatives [5,6] even if severe technical flaws have been highlighted 7]. NMEP consist of a prominent early biphasic component folowed by small amplitude polyphasic components. Based on data rovided by collision studies [8,9], it appears of particular imporance during IOM performed with NMEP to look at the persistence f polyphasic components. Some true positive NMEP alerts are omposed of a disappearance of this polyphasic component with ersistence of the earlier monophasic component [10]. In this bservation (No 4), a wake up test elicited satisfactory movement in oth legs. Instrumentation was thus maintained. This patient postperatively presented brisk reflexes in both legs lasting 10 days, ithout sensory modification, and then recovered completely. oreover, we have the experience of NMEP alerts without signifcant change of SSEP. As an example, a patient post-operatively resented a proximal motor deficit with altered MEP and preerved SSEP, MRI showing increased T2 signal in the lateral part f the spinal cord. Based on these collision studies and clinical ata, it seems thus rather more logical to use the nomenclature