Safety and performance of the Spectra Optia apheresis system for white blood cell depletion in patients with elevated white blood cell counts

For the past 30 years, white blood cell depletion (WBCD) or leukocytapheresis has been conducted to rapidly reduce excessive circulating white blood cell (WBC) concentrations in patients at risk for or with symptoms of leukostasis due to hyperleukocytosis. The goal of leukocytapheresis is to prevent or treat acute complications from leukostasis, thereby enabling patients to receive potentially curative chemotherapy.


| INTRODUCTION
For the past 30 years, white blood cell depletion (WBCD) or leukocytapheresis has been conducted to rapidly reduce excessive circulating white blood cell (WBC) concentrations. Candidates for leukocytapheresis are patients at risk of, or who have signs and symptoms of leukostasis due to hyperleukocytosis. Hyperleukocytosis is defined as a WBC count in excess of 100 Â 10 9 cells/L in patients affected by acute leukemia and is often associated with increased morbidity and mortality. 1,2 Hyperleukocytosis incidence rate varies by disease type but it has been observed in an estimated 5%-20% of patients with untreated acute myeloid leukemia (AML) as well as in patients with acute lymphoblastic leukemia (ALL), chronic myeloid leukemia (CML) and chronic lymphocytic leukemia (CLL). 3 Clinically, hyperleukocytosis may cause leukostasis, resulting in impaired blood flow in the microcirculation of the central and peripheral nervous system, which presents clinically as headache, dizziness, vertigo, nystagmus, hearing loss, visual impairment, and somnolence, with coma and seizures occurring in severe cases. [4][5][6] Symptomatic leukostasis is often observed in the lungs, central nervous system, and kidneys of patients with acute leukemia at the time of initial diagnosis or relapse, and is associated with 2-to 3-fold increase in early mortality in both adults and children. [5][6][7] The American Society for Apheresis (ASFA) guidelines recognize leukocytapheresis as a first-line therapy for symptomatic leukostasis in acute forms of leukemia. 8 The clinical role of therapeutic leukocytapheresis in patients with AML with hyperleukocytosis or in patients with symptoms of leukostasis has been previously established. 5,[9][10][11] Briefly, patients with leukocytosis manifestations often cannot undergo definitive chemotherapy due to their rapidly declining health status. Therefore, the goal of leukocytapheresis in these patients is to help reduce the WBC burden thereby reducing acute complications and enabling these patients to receive the necessary potentially curative high-dose chemotherapy. 12,13 Leukocytapheresis along with medical management can mitigate many of the complications resulting from leukocytosis, but per ASFA guidance, definitive cytoreductive therapy should not be delayed in order to perform a WBCD procedure.
This report details the results from two clinical studies, one conducted in the European Union (EU) and one in the People's Republic of China, which assessed the use of the Spectra Optia Apheresis System (here-in referred to as Spectra Optia) for leukocytapheresis in patients with hyperleukocytosis. The primary objective of both studies was to assess if the WBCD procedure could safely and efficiently remove WBCs in patients with elevated WBC counts.

| Study conduct
This report outlines the results from two multicenter, single-arm studies which assessed the use of the WBCD procedure on the Spectra Optia (Terumo BCT, Lakewood, CO, USA) in patients with elevated (≥ 100 Â 10 9 cells/L) WBC counts. The first study (NCT02302365) was a retrospective data collection study to evaluate the performance and safety of the WBCD procedure over a pre-specified 2-year period at three European sites (UZ Leuven in Belgium, Institute for Transfusion Medicine and Immunohaematology in Frankfurt Germany, and Saint Istv an and Saint L aszl o Hospital in Hungary). The second study was a prospective, single arm, single procedure trial to characterize the performance and safety of the WBCD procedure in patients with a WBC count ≥100 Â 10 9 cells/L at eight sites in the People' Device performance and safety data were collected both prospectively and retrospectively. Data collected pre-procedure included medical history, demographics, diagnoses, concomitant medications, symptoms, vital signs, and laboratory measurements (eg, complete blood count [CBC]). During and following each procedure, data were collected for basic procedural and device information, clinical laboratory measurements from both the participant and the depletion product (waste bag), if available, blood products transfusion, medications, or replacement fluids required, adverse events (AEs) as defined by International Organization for Standardization (ISO) 14 155 Section 3.2, and device deficiencies.
The study protocols were approved by ethics committees (EC) at all sites and adhered to the principles that have their origins in the Declaration of Helsinki, the International Conference on Harmonization Guidelines for Good Clinical Practice, and ISO 14155:2011(E). Written informed consent was waived by local ECs for the retrospective study as no participant-identifying data was collected. For the prospective study, all participants signed the ECapproved informed consent form prior to any study procedure being conducted.

| Selection and description of participants
For the retrospective study, medical records between November 2011 and December 2013 from 3 sites in the EU were screened. Patients were eligible if they had received a minimum of one routine-use WBCD procedure on the Spectra Optia and had both pre-and post-procedure WBC count measurements (primary endpoint). As this was a routine-use study, pre-established parameters for patient selection (ie, adults, WBC count ≥100 Â 10 9 /L, etc.) was not conducted.
For the prospective study, adult participants (18 years or older) with a WBC count ≥100 Â 10 9 /L, a platelet count ≥30 Â 10 9 /L and adequate venous access were eligible to enroll. Participants must not have (1) had any significant bleeding in the 24 h prior to enrollment as determined by the physician, (2) undergone a WBCD procedure in the previous 7 days, (3) had a known hypersensitivity or condition that prevented the use of anticoagulants, and (4) female participants could not be pregnant or lactating. Patients diagnosed with acute promyelocytic leukemia (AML-M3) could not be enrolled in the study as leukocytapheresis is contraindicated in these patients due to the risk of accompanying disseminated intravascular coagulation, which is characteristic for this subtype of AML. 12 Patient demographics and study endpoint data are presented for each study individually and combined as patients in both studies underwent a leukocytapheresis procedure on the Spectra Optia due to hyperleukocytosis and are therefore representative of the patient population of interest.

| Procedure characteristics
The Spectra Optia System was used for all WBCD procedures and was comprised of (1) hardware which consists of the actual apheresis machine and an associated removable centrifuge filler, (2) a sterile single-use disposable tubing set (the Spectra Optia IDL set), and (3) embedded software for the WBCD procedure. In the retrospective study software Version 7.2 (or higher) was utilized at all sites while in the prospective study software Version 11.3 was utilized.
For the prospective study, the run target was set at a minimum of 2 times total blood volume (TBV), peripheral access was required, and the only anticoagulant allowed was Anticoagulant Citrate Dextrose -Solution A (ACD-A). A procedure was considered unsuccessful if the volume of whole blood processed was less than 2 Â TBVs. Replacement fluid, if required, was administered per standard of care in both studies. As the retrospective study was a routine use data collection study, the target run, type of anticoagulant, type of venous access, and the type of replacement fluid used (if required) was collected if present in the medical records. All identified WBCD procedures were included in the retrospective data set if they had a WBC pre-and post-count.

| Clinical endpoints
Study endpoints included the percent decrease in WBC count in the blood following a WBCD procedure, the percent of processed WBCs collected (ie, collection efficiency [CE1] for WBCs), and AEs observed within 24 h post-procedure. Adverse events were defined per Section 3.2 of ISO 14155 as any untoward medical occurrence, unintended disease or injury, or untoward clinical signs (including abnormal laboratory findings) in participants, users or other persons, whether or not related to the investigational medical device and whether anticipated or unanticipated. Adverse events were graded using the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE), Version 4.03, and coded using the Medical Dictionary for Regulatory Activities (MedDRA, Version 17.0 and 21.1). Device deficiencies were defined per Section 3.19 of ISO 14155:2011 as an inadequacy of a medical device with respect to its identity, quality, durability, reliability, safety or performance, and includes malfunctions, use errors and inadequate labeling.
The calculation for CE1 (see below) incorporated preand post-apheresis WBC counts thereby accounting for WBC removal via the procedure and WBC mobilization during the procedure. 14 For both studies, the apheresis parameters were determined according to the Spectra Optia Operator's Manual and site Standard Operating Procedures (SOPs) for each site. The collect flow was calculated by the Spectra Optia software based on the preprocedure WBC count and the Inlet Flow Rate in order to efficiently collect the large number of target WBCs at the collect port. The initial Collection Preference was set at a default at the start of the collection procedure and operators were instructed to increase or decrease the Collection Preference to achieve a desired target collection (based on collect line colorgram or hematocrit [if being tracked during procedure]) as necessary.

| Statistics
A sample size of 20 procedures was prespecified for the retrospective study and up to 68 participants (to acquire 53 evaluable procedures) for the prospective study. The sample size for the retrospective study was based on the expected number of procedures to be conducted at 3 sites over 2 years. The sample size for the prospective study was based on the primary endpoint, a preestablished objective performance criterion (OPC) for mean reduction of WBCs after a WBCD procedure. The OPC was calculated using data from the retrospective study along with data from a systematic literature review. The OPC/primary endpoint for the prospective study was an OPC of ≥46% mean reduction of WBCs.
Statistical analyses were performed using SAS (SAS Institute, Cary, NC, USA) software Version 9. Primary analysis of efficacy was conducted using the Full Analysis Set (FAS), which consisted of participants who completed the WBCD procedure with all primary endpoint measurements. Efficacy endpoints were calculated as follows: Percent decrease in WBC count (primary endpoint) was calculated as: CE1 was calculated as: Â Volume ml blood processed ð Þ Safety data were collected using the Safety Set (SS) which was comprised of all participants who initiated a WBCD procedure.  Estimated by Nadler's formula for total blood volume based on gender, height, and weight. b Documented for each procedure; participants may be included in >1 diagnosis/indication category.

| Participant disposition and demographics
The SS consisted of 75 participants who completed 90 procedures. The FAS consisted of 72 participants completing 87 WBCD procedures. Three participants were excluded from the FAS (all in the prospective study) due to an unsuccessful completion of the WBCD procedure (< 2 Â TBV processed). Data was retrospectively collected from 43 participants who underwent 58 WBCD procedures and prospectively from 29 participants who underwent 29 WBCD procedures. For the prospective study, data monitoring was continuously performed and the trial was stopped early after enrolling 32 patients due to the study not meeting the primary endpoint (OPC of ≥46% mean reduction of WBCs) Please see discussion for more details.

| Procedure characteristics
The WBCD procedure characteristics are presented in Table 2. As outlined in Table 2, the average duration of the WBCD procedures was 226 ± 83.2 minutes with a maximum duration of 581 minutes. The majority (n = 56) of procedures used ACD-A as the anticoagulant with an average volume of 1012 ± 459.5 mL per procedure. Fifty-four (54) procedures used a replacement fluid, with saline (n = 29) being the most common replacement fluid used. There was an average of 10 345.8 ± 3662.11 mL of whole blood processed during each procedure and 1389.2 ± 858.51 mL of fluid collected in the waste bag. Table 3 shows that the pre-procedure WBC counts were expectedly high for all participants with a mean WBC count of 237.7 Â 10 9 /L (range 67-895 Â 10 9 /L). The mean percent change in participant WBC counts postprocedure was 50.3 ± 21.2%, with a maximum decrease of 87%. The CE1 of the WBCD procedures was 53.7 ± 19.77%. One site in the retrospective study (UZ Leuven in Belgium) did not collect the WBC counts from the waste bag contents; therefore, this site did not contribute data to the CE analysis. Assessment of other hematological parameters preand post-procedure demonstrated the percent decrease of hematocrit and platelet count after a WBCD procedure were 7.5 ± 13.85% and 22.7 ± 49.55%, respectively.

| Device performance
A breakdown of the mean percent reduction and CE1 by diagnosis is presented in Table 4. Post-hoc analysis by diagnosis showed a significant (P < 0.001) difference in the CE1 between CML patients and AML, ALL and CLL patients. When the CE1 for other diagnoses was compared to the CE1 of CML, there was a trend of difference, but it did not reach statistical significance (P = 0.058). There was no difference observed when assessing the CE1 of AML, ALL and CLL patients to each other.

| Adverse events
Safety was assessed from the start of the procedure until 24 h post-procedure. One site in the EU (the Institute for Transfusion Medicine and Immunohaematology in Germany) did not collect 24 h AE data and was not included in the analysis. Between the 2 trials, there was a total of 61 participants (95.3%) who experienced a total of 279 AEs (Table 5). The most common AEs included thrombocytopenia (56.3%), hypoalbuminaemia (32.8%), hypokalemia (29.7%), anemia (26.6%), hypocalcemia (23.4%) and prothrombin time (PTT) prolongation (10.9%). True laboratory defined hypocalcemia was not evaluated as ionized calcium was not measured. Instead, the Preferred Term of hypocalcemia was used when patients reported clinical signs of hypocalcemia such as tingling, metallic taste in mouth, nausea, etc. One AE (nausea during rinseback) was determined to be related to the device and procedure while 31 AEs (48.4%) were related to the WBCD procedure and 54 AEs (84.4%) were related to the underlying disease. All-causality AEs reported for >10% of participants were primarily related to post-procedure changes in hematology or serum chemistry, which were either outside the normal reference range based on validated reference intervals or had a clinically significant change post-procedure when compared to pre-procedure values. During the study, 26 patients (40.6%) required a RBC transfusion, 18 patients (28.1%) required a platelet transfusion and 10 patients (15.6%) required a plasma transfusion. The total number of transfusions includes transfusions given pre-procedure, during the procedure and within 24 h post-procedure. These types of changes and the resulting transfusion(s) to treat the decreased laboratory values was anticipated within this very sick leukemia patient population.
The reported severity of all-causality AEs was 7.8% Grade 1, 31.3% Grade 2, 39.1% Grade 3, 21.9% Grade 4 and 1.9% Grade 5. There were 2 serious adverse events (SAEs) reported, respiratory failure (n = 1) in the retrospective study and cerebral hemorrhage (n = 1) in the prospective study. Neither SAE was related to the Spectra Optia device. Additionally, there was 1 participant in the prospective study (same participant with the SAE) that reported an unanticipated adverse device effect (UADE). This participant was found to have an increase in brain natriuretic peptide. Two participants (2.3%), one in each study, terminated the WBCD procedure prematurely. In the retrospective study, the participant discontinued due to a vascular access complication, while in the prospective study one participant reported 4 AEs (cardiac failure, hypoalbuminemia, blood lactate dehydrogenase increase and platelet count decrease) that led to study discontinuation. Four device deficiencies were reported but none led to study withdrawal or an AE/SAE.

| DISCUSSION
This multi-study report describes the performance and safety of the WBCD procedure conducted on the Spectra Optia Apheresis System in patients with hyperleukocytosis. This report provides individual and combined data from a retrospective study conducted at 3 sites in the EU and a prospective trial conducted at 8 sites in the People's Republic of China. In both studies, participants who had elevated WBC counts underwent at least 1 WBCD procedure on the Spectra Optia and the safety and performance of the WBCD procedure were assessed.
In the prospective study, each patient underwent 1 WBCD procedure while in the retrospective study, 13 of the 43 subjects had more than 1 WBCD procedure. Of these 13 patients, 11 underwent 2 procedures and 2 had 3 procedures with each additional procedure performed on consecutive days. Nine patients who had a second or third procedure did so to treat leukocytosis while 4 were to treated increased blood viscosity and/or tumor lysis syndrome. The purpose of these additional WBCD procedures was to lower the WBC burden thereby decreasing these clinical manifestations and stabilizing the patient to undergo additional therapy.
The mean CE1, which accounts for WBC removal via the procedure and WBC mobilization during the procedure, collectively was 53.7% and individually was 58.7% and 46.3% in the retrospective and prospective studies, respectively. There was an overall 50.3% decrease in the WBC count following the depletion procedure with a 54.5% decrease and 41.9% decrease in the retrospective and prospective study, respectively. The overall percent reduction of circulating WBCs did not meet the preestablished OPC in the prospective study and was observed to be less in the prospective study than in the retrospective study potentially due to the higher number of CML patients (n = 12 of 32 total patients) enrolled in the prospective study. Ad hoc analysis by diagnosis (Table 4) showed all CML patients had lower CE1 and percent decrease (27.7% and 30.0%, respectively) when compared to AML patients (60.2% and 54.1%). The analysis of variance (Table 4) shows there was a significant difference in CE1 between CML patients and AML, CLL, ALL patients indicating the WBCD procedure is not as effective in this patient population. This observation has been noted in other studies assessing WBCD in leukemia patients, most notably Nguyen et al who found that the WBC removal efficiency was dependent on diagnosis and was 71%, 66%, and 39% for ALL, AML, and CML, respectively. 13 The leukemic cells in CML patients are primarily mature leukocytes which are denser than the large myeloblasts found in AML patients. 15 These mature myeloid cells are harder to separate from the RBC layer during the WBCD procedure which can explain the lower reduction of circulating WBCs in the CML patients. Due to these limitations, leukapheresis is often not part of the standard of care for CML patients. The evidence from our study, ASFA guidelines, and the Nguyen et al study, suggest a single WBC depletion procedure in CML patients may not be enough to reduce the high levels of circulating WBCs. Instead, the WBCD procedure may be beneficial in AML and ALL patients to reduce the WBC burden prior to initiation of cytoreductive therapies.
Data collected in this study were gathered from patients suffering from hyperleukocytosis. These patients are often very sick, manifesting anemia, thrombocytopenia, and organ disease due to their underlying disease process. Hyperleukocytosis clinical symptoms result from large numbers of circulating WBCs due to rapid WBC proliferation and turnover as a result of primary bone marrow disease with abnormal hematopoietic blood cell production. 16 The high level of circulating WBCs result in patients presenting with clinical symptoms of anemia, thrombocytopenia and hyperleukocytosis which can cause hyperviscous blood and leukostasis (ie, microvessel occlusions). 16 The frequency and type of AEs observed during the WBCD procedures and within 24 h post-procedure along with the number of blood transfusion required pre-procedure through 24 h post-procedure reflect the baseline status of this sick leukemic patient population. There were no observed safety signals associated with the Spectra Optia device. The majority of AEs (84.4%) were considered related to the participants' underlying disease while approximately half were related to the WBCD procedure. These included hematologic and blood chemistry abnormalities (eg, low platelet count, anemia, and hypocalcemia) and are representative of the participants' underlying leukemic disease and/or the apheresis procedure (eg, hemodilution and anticipated platelet loss).
It is expected during an apheresis procedure for patients to lose a small and non-clinically significant number of platelets and RBCs due to the presence of these cellular components at the WBC/buffy coat collection interface. As the centrifugation process cannot completely separate WBCs from RBCs and platelets, some of these blood components are collected into the waste bag. Acute leukemia patients often have anemia and thrombocytopenia at presentation and the effects of leukocytapheresis on these conditions should be monitored. As expected in this patient population, preprocedure and post-procedure hematocrit and platelet counts were observed to be low for the majority of participants. Spectra Optia has the ability to deliver a variety of replacement fluids to meet specific patient transfusion requirements as clinical teams see fit and the majority (>60%) of patients received a blood component transfusion and/or replacement fluid. Even though, the WBCD procedure further reduced the platelet and hematocrit levels the extent of depletion (7.52% and 22.74%, respectively) was relatively small and if needed could be managed through RBC or platelet transfusions. 4,5,17 There were only 2 SAEs, one from each study, and neither were related to the Spectra Optia.

| CONCLUSIONS
The data collected within these studies indicate that WBCD procedure using the Spectra Optia is well tolerated and safe in a critically ill patient population, as evaluated by percent decrease in WBC count, CE1 for WBC, and AE incidence.

PATIENT CONSENT
Written informed consent was waived by local Ethics Committees for the European retrospective study as no participant identifying data was collected. For the prospective study, all participants signed the Ethics Committee-approved informed consent form prior to any study procedure being conducted.

CLINICAL TRIAL REGISTRATION
The retrospective data collection study was registered on Clinicaltrials.gov under NCT02302365. The prospective study was not registered as it did not meet the criteria for registration per Applicable Clinical Trial (ACT) Under 42 CFR 11.22(b).