A comparison of chronic manual and automated red blood cell exchange transfusion in sickle cell disease patients

Chronic transfusion is one of the few effective means for treating or preventing the multisystem complications in sickle cell disease (SCD). Red blood cell (RBC) exchange transfusion (RBCX) is often preferred over simple transfusion because it rapidly reduces sickle haemoglobin (HbS) without raising final haematocrit (Hct), and there is some evidence that it may further reduce the risk of subsequent stroke (Hulbert et al, 2006). With new indications emerging (Casella et al, 2010), it is imperative that effective means of chronic transfusion to maintain appropriate haematological and clinical targets are identified. Some centres advocate the use of manual RBCX, where patients undergo sequential phlebotomies and transfusions at each session as it is a less resource-intensive procedure than automated RBCX. Proponents of automated RBCX argue that it is achieves haematological targets quickly and more consistently. However, no observational or controlled studies comparing automated with manual RBCX in either the paediatric or the adult setting have been reported. This study examined whether adult SCD patients on manual RBCX differ from those on automated RBCX in their ability to achieve pre-defined haematological targets, rate of complications, blood usage and clinical outcome over a 1-year period. This retrospective observational cohort study was conducted as part of a quality assurance audit between 1 May, 2011 and 30 April, 2012 at two SCD comprehensive care centres in London, United Kingdom. Subjects were included if they had a confirmed diagnosis of SCD, received regular RBCX (>1 session) between 1 May 2011 and 30 April 2012 and were aged >16 years. Bart's Health NHS Trust (BHT) performs manual RBCX while Guy's and St. Thomas NHS Trust (GSTT) performs automated RBCX exclusively. Both institutions adhere to the same clinical standards of comprehensive care (Olujohungbe, 2008), indications for chronic transfusion and pre-RBCX HbS targets (Table SI). All RBC units are provided by the UK National Blood Service [average packed RBC Hct (pRBC) 0·6]0, leucodepleted, HbS negative, ≤7 d old, phenotypically matched (for D, C/c, E/e, Kell antigens) and antigen-negative for any known alloantibodies. A total of 51 subjects, totalling 401 RBCX sessions and 38·8 patient-years of observation, were included. Table 1 summarizes the patient demographics and baseline characteristics. The distribution of the pre-RBCX HbS/SC fraction was found to be highly variable both within and between subjects (Fig 1). Although the mean pre-RBCX HbS/SC fraction was not significantly different between the two cohorts (P = 0·212), a greater proportion of subjects in the automated RBCX cohort were able to consistently achieve the pre-RBCX HbS/SC target compared to the manual RBCX cohort (P = 0·048). After adjusting for the prescribed pre-RBCX HbS/SC target, the Mantel-Haenszel Odds Ratio (OR) estimate for not attaining target because manual RBCX was employed was 4·72 [95% confidence interval (CI): 0·89–25·20]. On multivariate logistic regression, only older age was associated with a higher likelihood of consistently achieving the pre-RBCX HbS/SC target (OR 1·12, P = 0·012). There was no significant difference in the median post-RBCX haematocrit (P = 0·931). Automated RBCX utilized more pRBC in both the volume (P < 0·0001) and units (P < 0·0001) but required half the time (P < 0·0001) and was performed less frequently (P = 0·001) than manual RBCX (Table 1). Treatment failure, defined as a SCD-related complication that the RBCX was intended to prevent, did not occur in any subjects within the defined observation period, with the exception of pain. No significant difference in adverse event rates was observed (Table 1, P = 0·7953). The majority of the adverse events in the manual RBCX cohort were termination of RBCX procedure due to poor intravenous access, blocked line or port, while the majority of the adverse events in the automated RBCX cohort were attributed to dizziness or hypotension. In this first systematic comparison of two RBCX methods, a higher proportion of the automated RBCX cohort was able to consistently achieve the prescribed pre-RBCX HbS/SC target while utilizing less time than the manual RBCX cohort. However, a large number of subjects in both cohorts were not able to achieve their prescribed target. Published literature has shown that the ability to achieve the requisite pre-RBCX HbS/SC target can be quite variable (Aygun et al, 2012; Ware & Helms, 2012). The evidence for the chosen pre-RBCX HbS/SC target of <30% was largely derived from the STOP (Stroke Prevention in Sickle Cell Anaemia) and STOP2 trials (Adams et al, 1998; Adams & Brambilla, 2005), the SWiTCH (Stroke With Transfusions Changing to Hydroxyurea) trial (Ware & Helms, 2012) and retrospective data (Pegelow et al, 1995). Indications without the support of randomized controlled trials (RCTs) were largely drawn from retrospective cohort studies (Cohen et al, 1992; Miller et al, 1992) or expert opinion. Performing a larger volume manual RBCX may improve the pre-RBCX HbS/SC but will present a formidable challenge as the duration of the RBCX session will be lengthened beyond 4 h and will increase the risk of hypotension and syncope. The major advantage of automated RBCX is that it can circumvent all of the above issues and optimization of outcomes can be achieved by shortening the RBCX interval. Patients with low baseline haematocrit and those who are not consistently achieving the prescribed pre-RBCX HbS target may be placed on automated RBCX as this can overcome the limitations imposed by manual RBCX. However, the technical and resource demands of automated RBCX may prevent its widespread adoption in resource-constrained institutions unless it can be shown to be more cost-effective than manual RBCX. Limitations of this study include its retrospective nature, the relatively short follow-up period, selection bias and unknown systematic differences between the two cohorts; these limitations could be resolved by performing a RCT. Even though a significant number of subjects were unable to consistently reach their prescribed pre-RBCX HbS/SC targets, the lack of neurological sequelae and treatment failure (with the exception of pain) suggests that a less stringent target may be sufficient. This question will need to be answered first by hypothesis-generating observational studies examining the longitudinal clinical outcomes in relation to their RBCX parameters, with the aim of proposing new HbS/SC targets. The proposed targets can then be compared to the current targets in a RCT setting. K.H.M.K.'s fellowship was funded in part by American Society of Hematology's Alternative Training Pathway Grant and an unrestricted educational grant from Novartis Canada. K.H.M.K. contributed to the concept, design and execution of the study, analysis and interpretation of data, critical writing and revising of intellectual content. R.W., J.H., and P.T. contributed to the concept and design of the study, interpretation of data, and revising of intellectual content. B.K. contributed to the interpretation of data and revising of intellectual content. All authors have given final approval for this version of the manuscript to be published. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.