Addition of plasma exchange to red cell exchange improves outcomes of fat embolism syndrome in sickle cell disease

Fat embolism syndrome (FES) is a severe but underdiagnosed complication of sickle cell disease (SCD) associated with very high mortality and morbidity. It is the result of extensive bone marrow necrosis that leads to release of large amounts of fat droplets in the systemic circulation. FES predominantly affects non-homozygous patients and those with a previously mild course of their illness. It is characterised by severe respiratory failure, neurological involvement, thrombocytopenia and extremely elevated levels of serum ferritin and lactiate dehydrogenase (LDH). Without intervention mortality exceeds 90% while immediate red cell exchange (RCE) substantially increases survival and it is currently the treatment of choice.1 Release of fat droplets in the circulation cause mechanical obstruction of the microvasculature. More than that, circulating phospholipids under the action of secretory phospholipase A2 (sPLA2) are metabolised to arachidonic acid which in turn acts as a substrate for the generation of a plethora of pro-inflammatory cytokines.2 Such mediators have a known nociceptive effect, enhance HbS polymerisation and further red cell sickling but also have the potential of causing direct tissue toxicity leading to end-organ damage.3, 4 Specifically in the brain, cytokines such as TNFα, IL-1, and IL-6 can initiate a positive feedback loop resulting in accumulation of extracellular glutamate which in turn causes great influx of water and intracellular oedema referred to as cytotoxic oedema. The corpus callosum and splenium are particularly susceptible to this effect due to a much higher concentration of cytokine receptors. Cytotoxic lesions of the corpus callosum (CLOCC) have been described in a plethora of conditions causing increased cytokine release such as drug reactions, infections and metabolic disorders5 while involvement of the splenium is invariably present in patients with cerebral FES.1 In addition to pro-inflammatory cytokines, SCD plasma contains several other harmful substances including adhesion molecules, free haemoglobin leading to nitric oxide depletion, arginase and procoagulant factors,6 while substances such as haptoglobin and hemopexin are depleted.7 Despite the increasing use of RCE in recent years, mortality from FES remains substantial while a large number of survivors are left with severe neurocognitive impairment (SNI).8 With that in mind, we have previously advocated addition of therapeutic plasma exchange (TPE) to RCE in order to address the inflammatory environment and potentially improve outcomes.8, 9 Here, we compare outcomes from FES using standard treatment with RCE and combination of RCE and TPE. From an ongoing systematic review of all FES cases ever published, we only selected cases published since 2012 to account for the advances in supportive care and only those where RCE was employed, and outcomes were compared to cases where a combination of RCE and TPE was used. We identified 39 cases treated with RCE alone since 2012. Mortality was 21%, 33% of patients were left with SNI and 13% mild neurological impairment (MNI) while only 33% made a complete recovery (CR). We are aware of 12 cases, published9-12 or in press, (separate from the previous 39), treated with RCE followed by TPE worldwide (eight UK, two Canada, one USA, one Denmark). Nine patients were adults and three patients were children. Six patients had homozygous Hb SS disease, five Hb SC and one Hb Sβ+. All patients, irrespective of genotype, previously had a mild course of their illness. All patients were treated in the intensive care unit and all received adequate and timely RCE to HbS/Hb S + C of 30% or less before plasma exchange. The number of TPE procedures performed varied greatly between cases (1–10) mostly depending on clinical response. There was one death (8%), two patients (17%) suffered MNI while eight patients (75%) achieved CR. It should be noted that there were no survivors left with SNI while the two patients reported as suffering MNI are patients who developed the complication recently, have fully recovered cognitive function but are still in rehabilitation improving from generalised weakness resulting from prolonged hospitalisation and are very likely to achieve CR eventually (Figure 1). Among the patients achieving CR there were patients with extensive brain involvement or respiratory failure requiring mechanical ventilation or even extracorporeal membrane oxygenation. There were no episodes of recurrence of FES among the patients treated with RCE and TPE with a mean follow-up of 26 months (4–108). Three patients had a CLOCC on brain magnetic resonance imaging of whom one was treated with RCE only and two with RCE + TPE. The fact that CLOCCs reflect direct cytokine-induced neurocytotoxicity lends further support to addressing the pro-inflammatory environment. The patient receiving RCE alone survived, but with very SNI, was wheelchair bound, and requiring 24-hr care in an appropriate facility. In contrast, both patients receiving the combination of red cell and plasma exchange had a very good clinical outcome with one achieving CR before discharge from hospital and the other currently recovering speedily with physiotherapy at home (Figure 2). In all cases, TPE was well tolerated with no associated adverse events. These early data suggest a clear benefit from this approach. Even though TPE can sometimes be associated with adverse events,13 it is overall a widely used intervention with a relatively safe profile. Therefore, we recommend its addition to RCE as standard management for FES in SCD. The duration of TPE should be tailored to clinical response as well as the trend in haematological and biochemical markers such as haemoglobin, platelets, serum ferritin and LDH. In these early cases we observed recovery of some very severely affected patients, therefore we recommend persevering with treatment even when signs of response are not immediately evident. Dimitris A. Tsitsikas designed the project and wrote the paper, Susan Rowe, Alessandra Bosch, Caitlyn Hui, Nandini Sadasivam, Nicolaos J. Palaskas, Shivan Pancham, Syed Rizvi, Joseph Taylor, Paul Greaves, Andreas Glenthøj, Marianne Hoffmann and Emma Drasar provided and analysed data and critically reviewed the manuscript and Perla Eleftheriou co-designed the project and critically reviewed the manuscript.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.