Pre-clinical experiment
The feasibility of extracting Plasmodium parasites from blood using apheresis was initially assessed using cultured P. falciparum parasites. The 1% HCT layer contained the greatest concentration of all parasites as determined by 18S qPCR, with a 1.3-fold concentration of all parasites and a 3.7 and 8-fold concentration for female and male gametocytes, compared to pre-apheresis (Additional File 6 Fig. 3 and Tables 1 to 3). The 2% HCT layer contained the greatest concentration of asexual parasites by microscopy (2.7-fold concentration; Additional File 6 Fig. 4A and Table 4). The highest concentration of gametocytes detected by microscopy was seen in the 1% HCT layer (76-fold ; Additional File 6 Fig. 4A and Table 4). Both asexual parasites and gametocytes were also visualised by microscopy in the 1%, 3% and 5% HCT layers (Additional File 6 Fig. 4B and Table 5). These results demonstrated the technical feasibility of the approach and the experiment was allowed to proceed to the clinical stage.
Clinical experiment
The course of P. vivax infection (Fig. 1) followed the same course as demonstrated in previous studies(9). Subjects 1 and 2 were treated with artemether/lumefantrine 10 days post malaria inoculation. To augment pre-apheresis parasitemia, treatment of subjects 3 and 4 was delayed to day 11. Apheresis could be delayed safely as the clinical signs observed for subjects 3 and 4 were mild enough for the artemether/lumefantrine treatment, only administered after apheresis, to be postponed by 24 hours. All subjects became 18S qPCR negative for parasites within 72 hours of treatment initiation.
A summary of the key differences in the planning of the clinical trial in each of the four subject cohorts is shown in Table 1. During cohort 4 a red cell depletion was carried out, producing an intermediate bag sample, followed by a second apheresis procedure on the red cell depletion product. The second apheresis procedure involved sampling of ~ 100 ml of the lowest HCT layers of the sample (final bag) followed by 100mls of the subsequent lowest HCT layers (spare bag) and then the remainder (waste bag) (Additional File 6 Fig. 2). A schematic of the sampling that occurred during cohort 4 can be seen in Additional File 6 Fig. 2.
Parasitemia as measured by 18S qPCR in all four subjects. Day 0 represents the day of inoculation. Apheresis occurred on day 10 for subjects 1 and 2, and day 11 for subjects 3 and 4.
Safety
A total of 68 AEs occurred in the 4 subjects (Table 2). No Serious AEs were reported. The majority of AEs were mild or moderate. Five severe AEs occurred in 4 subjects: one episode of neutropenia (0.68 × 109/L [0.45 × LLN]; duration 8 days), two of lymphopenia (0.43 × 109/L [ 0.43 × LLN] and 0.33 × 109/L [0.33 × LLN]; both lasting 3 days), and two of fever (both 40.2 °C; duration 30 minutes and 25 minutes). All severe AEs were transient and resolved by the end of the study. The majority of AEs (54/68; 79.4%) were attributed to malaria, while 9/68 (13.2%) were attributed to apheresis. These included neutropenia (one subject nadir 0.68 × 109/L) which was recorded three times due to changes in severity, two cases of lymphopenia (nadir 0.33 × 109/L and 0.75 × 109/L) and two cases of leukopenia (nadir 2.2 × 109/L and 2.4 × 109/L). One subject had an episode of herpes labialis (herpes simplex virus-1 PCR positive). One subject experienced mild hypophosphataemia (0.70 mmol/L) of two days duration. All AEs attributed to apheresis apart from the case of hypophosphataemia were considered to be possibly related to apheresis, malaria or a combination of both.
Table 2. Summary of the main safety findings
|
Subject 1
|
Subject 2
|
Subject 3
|
Subject 4
|
Total AEs
|
SAEs
|
0
|
0
|
0
|
0
|
0
|
AEs
|
20
|
13
|
15
|
20
|
68
|
AEs related to malaria
|
13
|
8
|
14
|
19
|
54
|
AEs related to apheresis
|
2
|
3
|
0
|
4
|
9
|
Max temp °C
|
40.2
|
38.8
|
40.2
|
39.6
|
N/A
|
Max malaria clinical score
|
8
|
1
|
2
|
7
|
N/A
|
Ibuprofen use
|
400 mg × 5
|
400 mg × 5
|
400 mg × 5
|
nil
|
N/A
|
Acetaminophen use
|
1 g × 4; 500 mg × 1
|
1 g × 2
|
1 g × 4
|
500 mg × 2; 1 g × 2
|
N/A
|
Peak ALT (IU/L)
|
111
|
118
|
47
|
80
|
N/A
|
Peak AST (IU/L)
|
83
|
57
|
42
|
44
|
N/A
|
Platelet Nadir (× 109/L)
|
119
|
98
|
99
|
75
|
N/A
|
Maximum drop in haemoglobin from baseline (g/L)
|
17
|
9
|
25
|
20
|
N/A
|
Lymphocyte Nadir (× 109/L)
|
0.33
|
0.43
|
0.58
|
0.75
|
N/A
|
Peak parasitemia (parasites/mL)
|
15943
|
35156
|
64243
|
44431
|
N/A
|
SAE; serious adverse event, AE; adverse event. Summary of the main safety findings encountered during the study.
Characteristics of samples collected by apheresis
The red blood cell counts of the various HCT layers were generally in alignment with what would be expected (Supplementary Table 6), except in two subjects. In subject 2 the red blood cell counts of the 2% HCT layer was closer to what would be expected from a 3% HCT layer and vice versa, and the 8% HCT sample in subject 3 had a HCT of 11%. The cell composition of samples collected using apheresis in subjects 1 to 3 (Supplementary Fig. 5A) showed an ~ 60 to 170-fold decrease in the RBC:WBC ratio from pre-apheresis samples compared to apheresis samples. Among samples collected by apheresis from subject 4, where a double apheresis process was undertaken, the RBC:WBC ratio was close to that of the pre-apheresis sample (Supplementary Fig. 5B), with the exception of the final bag sample (3% HCT). Reticulocyte counts measured on the Sysmex analyser were the highest in subject 2: 0.23 × 109 /L (reference range for whole blood: 25–120 × 109/L).
Concentration of asexual parasites
18S qPCR targets the highly conserved plasmodium 18S ribosomal RNA gene present in asexual parasites and gametocytes [14, 18]. However, based on the P. vivax life cycle, microscopy and P. vivax female gametocyte qRT PCR data (pvs25) it was determined that the vast majority of parasites, detected using 18S qPCR, were asexual parasites.
No single HCT layer contained > 40% of all the recovered P. vivax asexual parasites (Fig. 2). An increase in parasite concentration occurred as HCT increased in subjects 1 to 3 (Fig. 2A), with some variation in relative enrichment of parasites in apheresis samples compared to pre-apheresis samples at any given HCT (Fig. 2A). The highest concentration achieved was a 4.9-fold increase in parasite density in the 7% HCT layer in subject 1 (Fig. 2A and Additional File 6 Table 7). There was no apparent enrichment of parasites when the procedure was modified to include a second apheresis process (subject 4; Fig. 2B).
Fold enrichment of parasites as determined by 18 qPCR in samples collected using apheresis compared to the pre-apheresis samples in subjects 1 to 3 (A) and subject 4 (B).
When parasite concentration was adjusted for RBC count, all apheresis samples collected from subjects 1 to 3 demonstrated enrichment for asexual parasites compared to pre-apheresis (Fig. 3A). In general, when parasite counts were corrected for RBC count, parasite enrichment was highest in the low HCT samples with the highest enrichment detected in the 0.5% HCT sample in subject 3 (Fig. 3A and Additional File 6 Table 7; 138-fold compared to pre-apheresis). The relative concentration of parasites from subject 4, where the second apheresis procedure was performed, was observed in the lowest HCT samples. In particular, the final bag (3% HCT) and the spare bag (5% HCT) samples had relative enrichment levels of 20- and 8-fold respectively compared to pre-apheresis (Fig. 3B and Additional File 6 Table 7).
Fold enrichment of parasites corrected for red blood cell counts as determined by 18S qPCR in samples collected using apheresis compared to the pre-apheresis sample in subjects 1, 2 and 3 (A) and subject 4 (B).
Concentration of P. vivax gametocytes
Analysis of the apheresis samples from subject 4, where the double apheresis process was undertaken, demonstrated an increase in the level of female gametocytes of 1.45-fold compared to pre-apheresis, as determined by levels of the gametocyte-specific transcript pvs25 qRT PCR (Fig. 4B and Additional File 6 Table 8).
Subjects 1, 2 and 3 demonstrated a reduction in the level of female gametocytes compared to pre-apheresis (Fig. 4A and Additional File 6 Table 8). In most cases the reduction was > 10-fold. Percoll concentration of whole blood taken pre-apheresis resulted in a significant enrichment (up to 45-fold in subject 3) of pvs25 compared to pre-apheresis samples not enriched with percoll (Fig. 4C and Additional File 6 Table 8).
Fold enrichment of female P. vivax gametocytes determined by pvs25/ml in samples collected using apheresis compared to the pre-apheresis sample in subjects 1, 2 and 3 (A) and subject 4 (B). Fold enrichment of female P. vivax gametocytes determined by pvs25/ml in samples collected using apheresis compared to the pre-apheresis sample and pre-apheresis samples enriched with percoll in subjects 1, 2 and 3 (C).
In subject 1 the percoll enrichment experiment failed due to technical issues resulting from suboptimal processing of the sample.
When gametocyte concentrations were corrected for red blood cell counts, enrichment levels were generally higher in the lower HCT samples (Fig. 5A and 5B; Additional File 6 Table 8). Relative enrichment of gametocytes in subjects 1, 2 and 3 was lower than for total parasites, with a maximum enrichment of 6.2-fold in the 0.5% HCT layer (Additional File 6 Table 8). In subjects 3 and 4 enrichment relative to red blood cell count was observed (Fig. 5A and 5B). In subject 4, the greatest enrichment of gametocyte transcripts in the double apheresis process was observed in the spare bag (5% HCT) with an enrichment of 6.1-fold (Fig. 5B and Additional File 6 Table 8).
Fold enrichment of female P. vivax gametocytes corrected for RBC counts determined by pvS25 collected using apheresis compared to the pre-apheresis sample in Subjects 1, 2 and 3 (A) and subject 4 (B).
Flow Cytometry
In subjects 2 and 3, it was observed that as the haematocrit decreased there was an increase in the percentage of WBCs (Fig. 6A and 6B). A small percentage of SYBR Green I + cells and CD45-, which could represent either parasitised RBCs or reticulocytes, were detected in samples from subjects 2 and 3, where 2–3% HCT pooled samples from subject 3 had the highest concentration (0.59%) of SYBR Green I + and CD45- (Fig. 6B). In subject 4, where CD71 antibody staining was used to identify reticulocytes, the highest levels of reticulocytes (CD71 + cells) were observed in the final bag (3% HCT) and the spare bag (5% HCT) (Fig. 6C). Microscopic analysis of these samples suggested that the reticulocyte population detected by flow cytometry consisted of uninfected reticulocytes. No parasitised RBCs or reticulocytes could be detected by flow cytometry.
Percentage of CD45-/SYBR + in samples obtained from subject 2 (A) and subject 3 (B). The percentage of (CD71+) reticulocytes from subject 4 (C).
Microscopy
The high concentration of WBCs in apheresis samples, meant many of the thick and thin blood smears were extremely difficult to read (Additional File 6 Tables 9 and 10).
Among samples where it was possible to read the films, parasite counts were low, and in alignment with the counts obtained by qPCR (Additional File 6 Tables 9 and 10). Notwithstanding the technical difficulties in reading the slides and the low parasite counts, no apparent concentration of parasites was observed in any of the apheresis samples compared to pre-apheresis. The 3% HCT sample in subject 3 contained 7 parasitised cells compared to 0 in the pre-apheresis sample and the 1% HCT sample in subject 1 demonstrated a 4-fold increase in the number of visualised parasites compared to pre-apheresis. In general, higher parasite numbers were seen in the lower HCT layers. The vast majority of parasitized RBCs contained ring form parasites, with trophozoites and gametocytes observed in samples from subjects 3 and 4 only (Additional File 6 Tables 9 and 10).
Mosquito transmission
Membrane feeding of pre-apheresis or post percoll enrichment samples were undertaken in all 4 subjects, but for logistical reasons membrane feeding on apheresis samples could only be undertaken in subjects 1 and 4 (Table 3).
None of the samples from subject 1 or 2 resulted in successful mosquito transmission (Table 3). In subject 3, the infection rate from the percoll-enriched sample was 5.8-fold higher than the infection rate from the non-enriched pre-apheresis sample (99% vs 17.2% [Table 3]). Likewise, in subject 4, the infection rate from the final bag (3% HCT) was 5.5-fold higher than the infection rate from either the pre-apheresis sample or the intermediate bag (64%HCT) (pre-apheresis: 3.6%, intermediate bag: 4%, final bag: 20% [Table 3]).
The final sample bag (3% HCT) in subject 4 was the only sample obtained using apheresis that demonstrated an increase in transmission over samples obtained pre-apheresis. Due to logistic issues the spare bag (5% HCT), which contained the greatest enrichment for gametocytes compared to pre-apheresis when adjusted for red cell counts in subject 4 (Fig. 5B) was not subject to membrane feeding.
Table 3. Mosquito infection rates following membrane feeding assays.
Sample
|
Subject 1
|
Subject 2
|
Subject 3
|
Subject 4
|
Pre-apheresis (with-Percoll enrichment)
|
Apheresis samples
|
Pre-apheresis (with-Percoll enrichment)
|
Pre-apheresis (without-Percoll enrichment)
|
Pre-apheresis (with-Percoll enrichment)
|
*Pre-apheresis (without- Percoll enrichment)
|
Apheresis samples
|
1% HCT
|
2% HCT
|
3% HCT
|
Intermediate (64% HCT)
|
Final (3% HCT)
|
Waste (42% HCT)
|
Feeding rate (No. mosquitos fed/No. total mosquitos [%])
|
127/133 [95.5%]
|
101/103 [98.1%]
|
103/104 [99%]
|
114/116 [98.3%]
|
75/77 [97.4%]
|
102/103 [99%]
|
107/107 [100%]
|
29/29 [100%]
|
28/29 [97%]
|
26/26 [100%]
|
30/30 [100%]
|
Mortality rate (No. dead mosquitos/No. total mosquitos [%])
|
23/133 [17.3%]
|
3/103 [2.9%]
|
6/104 [5.8%]
|
11/116 [9.5%]
|
7/77 [9.1%]
|
6/103 [5.8%]
|
5/107 [4.7%]
|
1/29 [3.4%]
|
3/29 [10.3%]
|
1/26 [3.8%]
|
1/30 [3.3%]
|
Infection rate (No. mosquitoes with oocysts/No. mosquitos tested [%])
|
0/110[0%]
|
0/105 [0%]
|
0/98 [0%]
|
0/105 [0%]
|
0/77 [0%]
|
16/93 [17.2%]
|
99/100 [99%]
|
1/28 [3.6%]
|
1/25 [4%]
|
5/25 [20%]
|
0/24 [0%]
|
Only subjects 1 and 4 involved the testing of samples collected using apheresis. For logistic reasons, MFA was not carried out in subject 4. Following consideration of gametocyte levels, DMFA was not carried out in subject 2. Reported parameters include feeding rate of mosquitoes, adult mosquito mortality rate after feeding and mosquito infection rate.
*MFA without percoll enrichment to save on blood volume draw for safety reasons.
Additional 18S qPCR testing
To investigate for possible accumulation of parasites in the magnets and tubing structures in the apheresis cassette, qPCR testing was carried out on three blood clots with diameters of up to 1 cm found near two magnets, and a silo like structure that formed part of the single use apheresis tubing and processing cassette from subject 3 (Additional File 6 Fig. 11). Clots were thoroughly homogenised and tested by 18S qPCR. The greatest enrichment observed relative to pre-apheresis whole blood was 1.4-fold in the apheresis cassette magnet 2 sample (Additional File 6 Fig. 12 and Table 11).
To investigate whether haemolysis may be taking place in the apheresed blood resulting in release of parasite DNA into the extracellular fluid, plasma from apheresis samples in subject 3 was collected by centrifugation of the sample, with the plasma subject to 18S qPCR testing. However, no significant accumulation of parasite DNA was detected in the plasma compared to pre-apheresis (Additional File 6 Fig. 12 and Table 11).