Comparative study between chronic automated red blood cell exchange and manual exchange transfusion in patients with sickle cell disease: A single center experience from Saudi Arabia

Abstract

BACKGROUND:

Red cell transfusion remains the gold standard in managing sickle cell disease (SCD) with severe complications. Offering red blood cell exchange (RBCX) either manual exchange transfusion (MET) or automated RBCX (aRBCX) can reduce the complications of chronic transfusion and maintain target Hb thresholds. This study audits the hospital experience of overseeing adult SCD patients treated with RBCX, both automated and manual, and compares the safety and efficacy.

MATERIALS AND METHODS:

This retrospective observational study was conducted as an audit for chronic RBCX for adult patients with SCD in 2015–2019 at King Saud University Medical City, Riyadh, Saudi Arabia.

RESULTS:

A total of 344 RBCX for 20 adult SCD patients who were enrolled in regular RBCX, (11/20) patients had regular aRBCX with a total of (157) sessions, and (9/20) patients had MET with a total of (187) sessions. The median level of HbS% post-aRBCX was significantly lower than MET (24.5.9% vs. 47.3%, P < 0.010). Patients on aRBCX had fewer sessions (5 vs. 7.5, P < 0.067) with better disease control. Although the median yearly pRBC units per patient for aRBCX was more than the double needed for MET (28.64 vs. 13.39, P < 0.010), the median ferritin level was 42 μg/L in aRBCX versus 983.7 μg/L in MET, P < 0.012.

CONCLUSION:

Compared to MET, aRBCX was more effective in reducing HbS, with fewer hospital visits and better disease control. Although more pRBCs were transfused, the ferritin level was better controlled in the aRBCX group without increasing alloimmunization risk.

Introduction

Sickle cell disease (SCD) is an inherited autosomal recessive hematological disorder with severe and life-threatening complications. It is commonly seen in sub-Saharan Africa, the Mediterranean basin, and Saudi Arabia. There are around 300,000 new cases globally each year.[1] In Saudi Arabia, SCD is endemic in the eastern and southern regions with a prevalence of 145/10,000 and 24/10,000 consecutively.[2] There are many challenges in controlling SCD in Saudi Arabia, especially with the prevalence of consanguineous marriages (57.7%), which increases the risk of genetic diseases. To some extent, premarital screening programs have helped decrease the prevalence of SCD, but it remains high.[3]

Clinical severity of SCD manifests as hemolytic anemia and cycles of microvascular vaso-occlusion leading to end-organ ischemia-reperfusion injury and infarction. Recurrent hemolysis and vaso-occlusion promote inflammation that leads to progressive small- and large-vessel vasculopathy.[4] Red blood cell transfusion and hydroxyurea have been the only disease-modifying agents for the decades. L-arginine and voxelotor are novel therapies that improve the SCD course.[5,6] The recent approval of crizanlizumab by the American Food and Drug Administration will provide significant additional support to the management of SCD[7,8,9] by reducing sickle cell-related pain crises. However, stem cell transplantation and gene therapy remain the only potentially curative therapy.[1,10]

Exchange transfusion is generally preferred over top-up transfusion, especially for patients intended to be on regular transfusions. The chronic exchange transfusions are performed either by manual method (manual exchange transfusion [MET]) or by automated red blood cell exchange (aRBCX). On the “good” side, it rapidly decreases the HbS level without increasing the hematocrit or causing iron overload.[11] MET has many benefits as it is a simple procedure in which patients undergo sequential phlebotomies and transfusions at each session; it is less resource-intensive and does not require specialized training compared to aRBCX. However, MET appears to be less efficient in controlling SCD-related events and iron overload.[12] The “bad” side starts from the cost, the load on the transfusion services to provide a robust blood supply of which blood units must meet certain criteria, place an appropriate line, and maintain patients’ compliance. While the “ugly” facts about SCD management are lack of other options for patients with this bad disease and the impact of this on morbidity and mortality, and transfusion, in general, has its drawbacks, including allergic reactions, transfusion-transmitted infections, volume overload, iron overload, and alloimmunization.

Few studies compare chronic manual and aRBCX.[13,14] There is a previous abstract in Saudi Arabia that compared aRBCX versus MET for the treatment of acute complications of SCD.[15] Our current study is focused on comparing aRBCX versus MET for prevention of recurrence of multiple acute complications of SCD, and to our knowledge, this is the first study to be conducted within Saudi Arabia. It explores whether adult SCD patients on regular MET differ from those on aRBCX in achieving predefined hematological targets, disease control, the number of hospital visits, blood utilization, iron overload, and alloimmunization in 5 years.

Materials and Methods

Patient and study design

This is a retrospective observational cohort study, which was conducted as part of a quality assurance audit for a regular red cell exchange program for adult patients with SCD in 2015–2019 at King Saud University Medical City, Riyadh, Saudi Arabia. Institutional Board Review approval was obtained from the hospital research committee under research project No. E.19.4123. Regular RBCX program is intended for primary or secondary prophylaxis of SCD complications. Disease severity, patient preference, and appropriate peripheral venous access were the main factors that dictated the type of RBCX. Baseline laboratory testing and extended red cell phenotyping, including (RH, KEL, MNS, Duffy, and Kidd) were performed for each patient. The procedures were done at the medical outpatient day care unit. Complete blood count, hemoglobin electrophoresis, liver function test, ferritin, calcium and electrolytes, blood group, and antibody screen were performed routinely before each RBCX. According to ABO, D, C/c, E/e, and KEL antigens, RBC units matched to the patient's phenotype were selected, leukoreduced, HbS negative, and if possible fresh (7–10 days). MET was done as sequential phlebotomies and transfusions. The transfused blood volume was about 10 ml/kg, with variable bleeding according to HbS% and Hb levels at the previous exchange, individually assessed by the physician in the light of experience. The spectra optia apheresis system was used for aRBCX. All procedures were performed using peripheral venous access. The procedure was computer guided: Patient characteristics (height, weight, pre-RBCX hematocrit, mean hematocrit of the packed RBC units, fraction of cells remaining, and desired final hematocrit of the patients) were entered by the operator before starting the aRBCx. For both procedures, laboratory targets were (1) to obtain a postexchange HbS% level <30% and to maintain it <60% before the next procedure and (2) to maintain a hematocrit 28%–30% for both MET and aRBCX. Data on laboratory values, duration of the procedure, intervals between procedures, and adverse events were documented for each exchange session.

Statistical analysis

Statistical analysis was performed on Statistical Package for the Social Sciences (SPSS) software version 23 Armonk, NY: IBM Corp. Descriptive statistics for the continuous variables are reported as mean/median ± standard deviation, and categorical variables are summarized as frequencies and percentages. P < 0.05 is considered statistically significant.

Results

Three hundred and forty-four RBCXs were done for 20 adult SCD patients who were enrolled in the regular RBCX program. Table 1 summarizes patients’ demographics, characteristics, and the variables used to compare the effectiveness of both procedures, (11/20) patients had regular aRBCX with a total of (157) sessions and (9/20) patients had regular MET with a total of (187) sessions. These patients were candidates for exchange transfusion program because of the following indications: secondary stroke prophylaxis in 25% (5/20), acute chest syndrome (ACS) 15% (3/20), recurrent vaso-occlusive crisis (VOC) in 25% (5/20), fertility preservation in 5% (1/20), pregnancy in 5% (1/20), and recurrent priapism in 25% (5/20).

Table 1.

Demographic characteristics of patients on automated red blood cell exchange and either manual and the variables used to compare the effectiveness of both methods

Characteristics	Automated (n=11), n (%)	Manual (n=9), n (%)	P
Gender
Male	10 (91)	6 (67)
Female	1 (9)	3 (33)
Age (years), mean (range) (SD)	31 (20-41) (7.058)	38 (27-57) (8.992)	0.152
Blood group
O+	5 (45)	6 (67)
A+	4 (36)	3 (33)
B+	2 (18)	0
RH phenotype*
R0r (cDe/cde)	2 (18)	3 (33)
R1r (CDe/cde)	5 (45)	2 (22)
R2r (cDE/cde)	1 (9)	0
R1R1 (CDe/CDe)	1 (9)	0
R1R2 (CDe/cDE)	2 (18)	3 (33)
NA	0	1 (11)
K phenotype
Positive	1 (9)	2 (22)
Negative	10 (91)	7 (78)
Genotype
SS	8 (72)	8 (89)
S/β0	3 (28)	1 (11)
Indications
Secondary stroke prophylaxis	2 (18)	3 (33)
Acute chest syndrome	0	3 (33)
Vaso-occlusive crisis	3 (27)	2 (22)
Fertility preservation	1 (9)	0
Pregnancy	1 (9)	0
Priapism	4 (36)	1 (11)
Total number of RBCX sessions	157	187
Duration of enrolment (years), median (range)	2.9 (1-5)	3.15 (1-4.25)
Sessions/patient/year, median (range) (SD)	5 (3-8.7) (2.32)	7.5 (3-10.4) (2.6)	0.067
Interval between sessions (days), median (range) (SD)	73 (41.9-140) [4.5]	48.5 (28-84) (2.4)	<0.010
pRBC utilization; units/patient/year, median (range) (SD)	28.64 (12-57) (14.7)	13.39 (5.2-18.5) (4.5)	<0.010
Pre-RBCX HbS %, median (range) (SD)	54.4 (50.7-70) (6.3)	56.6 (32-74.5) (12.7)	0.331
Rate of target pre-RBCX HbS% maintained (<60%) between sessions (SD)	62.72 (22.3)	50.4 (28.5)	0.295
Post-RBCX HbS%, median (range) (SD)	24.5 (17-27.5) (9.4)	47.3 (41-68.5) (15.07)	<0.010
Rate of target post-RBCX HbS (<30%) (SD)	67.18 (18.77)	7.44 (17.516)	<0.010
Ferritin, µg/L, median (range) (SD)	42 (15.6-1215.6) (373.5)	983.7 (27.6-3098.3) (2317.4)	0.012
Iron chelation
Deferasirox	1	2
Deferoxamine	0	3
None	10	4
Alloimmunization	0	0
Transfusion reactions**	1	0
SCD crisis during the period of exchange†
Vaso-occlusive crisis (generalized body ache)	2	5
Acute chest syndrome	0	2
Priapism	0	1

*The most probable genotype was determined based on the most common Rh genes in the Caucasian population, **The patient developed a nonsevere allergic reaction in the form of mild rashes and pruritus, ^†Indicates new incidence of sickle cell disease crisis which required emergent aRBCX; some patients on MET developed more than one acute episode during the period of chronic RBCX. Mean HbS at the crisis was 74.07 (range 58.7-91.9). RBCX=Red blood cell exchange, pRBC=Packed red blood cells, SCD=Sickle cell disease, SD=Standard deviation, MET=Either manual, aRBCX=Automated red blood cell exchange

The median postexchange HbS% was (24.5% vs. 47.4%, P < 0.010) in aRBCX versus MET. In aRBCX, the HbS% was reduced to <30% in 67.18% (105/157) of the sessions. It was maintained below 60% between hospital visits in 62.72% (97/157) of the sessions with a median preexchange HbS% of 54.4%. However, in the MET group, the HbS% was reduced to <30% in only 7.44% of sessions. It was maintained below 60% between hospital visits in 50.4% (95/187) of the hospital visits with a median preexchange HbS% of 56.6%. Patients on aRBCX had fewer exchange sessions than MET (5 vs. 7.5, P < 0.067) in aRBCX versus MET. The median period between sessions was (73 days in aRBCX vs. 48.5 days in MET, P < 0.010). Most of the patients on aRBCX (9/11) remained stable without any complications except (2/11) who had one episode of VOC each during their follow-up period and were less compliant to the red cell exchange program than when compared to other patients. However, (5/9) of the patients on MET presented with a variable spectrum of SCD-related events [Table 1]. Although the average yearly pRBC units per patient needed for aRBCX was more than the double that was needed for MET (28.64 vs. 13.39, P < 0.010), the median ferritin level was 42 μg/L in aRBCX versus 983.7 μg/L in MET, P < 0.012. Only one patient of the aRBCX group needed iron chelation therapy for a short period compared to (5/9) in the MET group [Figure 1].

Figure 1.

Ferritin level in patients on regular RBCX. aRBCX: Automated red blood cells exchange, MET: Manual exchange transfusion; Y-axis demonstrate the serum ferritin level (ug/L) among each patient, X-axis demonstrates the session in which the ferritin was measured among the two groups. Each vertical line represents a different patient with his recorded ferritin readings throughout the follow-up period; 11 patients were on aRBCX, while 9 patients were on MET. Chelation therapy began at ferritin level ≥1000 ug/L

Out of the 344 procedures, only one transfusion-related adverse event in the form of mild allergic reaction was reported with rash and pruritus. None of the patients developed red cell allo-or autoantibodies during the follow-up.

Discussion

Chronic blood transfusion with simple or exchange transfusion is considered the cornerstone of comprehensive care for SCD patients.[16] It aims to increase the oxygen-carrying capacity by improving anemia, reducing the level of HbS present in the circulation, and preventing related systematic vasculopathy. It also includes monitoring and minimizing undesirable outcomes related to multiple transfusions as iron overload and alloimmunization. For primary and secondary stroke prevention, the recommendations in chronically transfused children with SCD to maintain HbS level below 30% between transfusions to avoid SCD-related events were primarily extracted from the STOP,[17] STOP2 trials (Stroke Prevention Trial in Sickle Cell Anemia)[18] and the SWiTCH trial (Stroke With Transfusions Changing to Hydroxyurea).[19] Higher HbS% before the next transfusion has been frequently reported in the literature (22%–84%), demonstrating the difficulty in complying with the recommendations.[13,14] As a start of the regular red cell exchange program in our institution, our hospital transfusion committee has chosen a less stringent target of 60% to be maintained to reduce costs and minimize the burden on the hospital transfusion services. Of interest, this practice in aRBCX cohort did not jeopardize disease control where only two incidents of VOC were reported in the observation period and were primarily attributed to the lack of patients’ compliance. However, recurrence of SCD-related complications was more than double in the MET group.

American Society for Apheresis (ASFA)[20] considers stroke prophylaxis as Category I for red cell exchange. Our experience showed that aRBCX and MET were equally effective in preventing the recurrence of neurological sequelae. For patients who were enrolled because of ACS, they were only in the MET arm, and only one out of three patients had a recurrence. Despite being ASFA uncategorized for SCD, nonacute indications, 25% of our SCD patients were enrolled in regular RBCX program to prevent recurrent priapism. No patients on aRBCX had a recurrence, whereas the MET patients have developed one episode of priapism. Ekong et al. have also reported the efficiency of regular aRBCX for preventing recurrence of priapism in a case series.[20]

A high proportion of aRBCX procedures achieved a prescribed postexchange HbS% target (67.18%), yet still, a significant number did not meet the target, mostly due to the unavailability of blood units with the required specifications. The rate of achieving a prescribed target was comparable to what has been reported in other regular exchange programs (70%).[21] In MET, only 7.44% achieved a target HbS%<30%. We could not compare this result to other studies, as other programs did not report post-HbS% in their papers. Our data also demonstrated no significant difference between the pre-RBCX HbS% for aRBCX and MET, as previously shown in other studies.[21,22,23] Nevertheless, more than 50% of SCD patients who were on MET developed different SCD complications. We think this is attributed mainly to the fact that the average intervals between MET procedures were inappropriately long (48.5 days; range, 28–84) when compared to other published data (28 days, range, 21–29 days)[22] and (35 days; range, 29–39 days).[23] In general, our experience shows that using aRBCX allows for longer intervals between two procedures with the absence of recurrent acute clinical events. The gain of time and stabilization of systemic organ involvement have contributed to decreasing the disease burden and fundamentally improved the quality of life.

In our RBCX policy, having an appropriate peripheral venous access was mandated for the chronic aRBCX to avoid central venous access. Indwelling central venous access has its complications such as infections, thrombosis, and difficulties in blood withdrawal,[24,25,26] and requires specialized care and may cause discomfort during daily life for the patient. Billard, et al. reported a good experience with a short-term femoral catheter insertion for each procedure,[27] and this choice might be considered in the future as an alternative for poorly controlled SCD patients who lack peripheral venous access.

With the use of aRBCX, the RBC requirement per procedure is significantly higher when compared to MET. However, ferritin level and iron overload were better controlled in the aRBCX cohort despite the higher number of pRBC utilization without chelation therapy for most of the SCD patients, which confirms the findings in previous studies.[22]

An additional concern for increased unit exposure in RBCX is alloimmunization. Our hospital's SCD protocol includes matching RBC units for RH and K antigens to minimize sensitization as recommended by the literature.[28,29] None of our patients developed alloantibodies in both cohorts. This becomes an interesting observation when we compare SCD patients on chronic RBCX to those who only received RBCX in an acute setting, where (10.42%) developed alloantibodies (unpublished data). One main reason is that many patients who come to our facility for the acute management of their SCD-related complications must have received transfusions at other hospitals where RH and K antigen matching is not their protocol. They also might have received blood at time of blood shortage when transfusion of antigen mismatched units was away better than no transfusion. This observation was previously reported in children with SCD where 24% of their chronically transfused patients developed alloantibodies before matching for Rh and Kell and reduced further when minor groups were also matched for.[30]

If a sickle cell care center provides an apheresis service, the cost does not seem to differ between aRBCX and MET when considering unit cost, hospital stay, and chelation cost as Dedekind's experience as well as other studies concluded.[22,23] However, we did not address the cost in our study.

After carefully evaluating our experience, we believe better outcomes for our RBCX program can be achieved by (1) shortening the intervals between the sessions, especially for MET, (2) re-evaluating and counseling patients who are poorly controlled on MET to be switched to regular aRBCX (3) adopting more strict parameters “postexchange HbS target <20%, and to be maintained below 50% until the next scheduled session” (4) enhancing patients’ compliance by health education and ameliorating their hospital experience and improving the quality of life.

Conclusion

Compared to MET, aRBCX is more efficient in reducing HbS% and sickle cell-related complications with longer intervals between hospital visits. Although more pRBCs were transfused, the ferritin level was better controlled in the aRBCX cohort without increasing alloimmunization risk.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.