BLOOD COMPONENTS Riboflavin and UV light treatment of platelets: a protective effect of platelet additive solution? Pieter F. van der Meer, Ido J. Bontekoe, Brunette B. Daal, and Dirk de Korte


Pathogen reduction technologies (PRTs) increase the safety of the blood supply, but are also associated with cell damage. Our aim was to investigate the effect of Mirasol PRT on platelet (PLT) concentrates stored in plasma and whether the use of a PLT additive solution (PAS) is able to improve in vitro quality. STUDY DESIGN AND METHODS: Twenty-two buffy coats (BCs) were pooled and split into two equal parts. To one half, 2 units of plasma were added, and to the other, 2 units of SSP1 PAS were added. Each part was equally split in half again (to resemble pooling five BCs) and PLT concentrates were prepared. One plasma PLT concentrate was Mirasol treated, and the other served as control; similarly, one SSP1 PLT concentrate was Mirasol treated, and the other not. PLT concentrates were stored for 8 days (n 5 12). RESULTS: Mirasol PRT led to elevated lactate production in PLT concentrates in plasma, giving lower pH values throughout storage. The use of SSP1 mostly abrogated this effect, and Mirasol-treated PLT concentrates in SSP1 had only slightly higher lactate production rates and annexin A5 binding as control PLT concentrates in plasma. However, irrespective whether plasma or SSP1 was used, Mirasol PRT led to higher CD62P expression and lower hypotonic shock response (HSR) scores. CONCLUSION: Mirasol treatment leads to higher PLT activation and lower HSR scores both when stored in plasma or SSP1. However, if Mirasol-treated PLTs are stored in SSP1, lactate metabolism and annexin A5 binding are lower, showing that PAS can partly mitigate the effect of PRT. The clinical relevance of this finding needs to be demonstrated.


athogen reduction of blood components is performed for various reasons: to further reduce the already small risk of viral transmission,1 to inactivate pathogens that are not tested for,2 to replace gamma irradiation,3 to reduce the occurrence of alloimmunization and thus to prevent platelet (PLT) refractoriness,4 and to reduce the risk of bacterial outgrowth.5 There are various methods for pathogen reduction technology (PRT) of plasma, PLT concentrates, red blood cells (RBCs), and whole blood, either commercially available or under active development. It is generally accepted that an increased safety can be associated with some loss of in vitro and in vivo quality. One such technology is the Mirasol PRT, which uses riboflavin in combination with UV illumination, and is commercially available for PRT of PLT concentrates and plasma in a number of countries. In vivo recovery of Mirasoltreated Trima apheresis PLTs suspended in plasma was 50.0 6 19.8% versus 66.5 6 13.4% for untreated PLTs and survival was 104 6 26 hours for treated and 142 6 26 hours for control PLTs, respectively (p < 0.001).6 Clinically, the 1-hour corrected count increments (CCIs) were 11.0 6 1.0 (mean 6 SE) for Mirasol-treated PLTs suspended in plasma and 16.6 6 1.0 for control PLTs (p < 0.0001), and the 24-hour CCI was 7.2 6 0.8 for treated and 10.1 6 0.8 for untreated PLTs suspended in plasma, respectively (p 5 0.0027).7

ABBREVIATIONS: ACE 5 accumulated centrifugal effect; BC(s) 5 buffy coat(s); HSR 5 hypotonic shock response; PRT(s) 5 pathogen reduction technology(-ies). From the Department of Product and Process Development, Sanquin Blood Bank, Amsterdam, the Netherlands Address reprint requests to: Pieter van der Meer, PhD, Department of Product and Process Development, Sanquin Blood Bank, Plesmanlaan 125, 1066 CX Amsterdam, PO Box 9713, 1006 AC Amsterdam, the Netherlands; e-mail: [email protected] Received for publication October 14, 2014; revision received January 13, 2015; and accepted January 13, 2015. doi:10.1111/trf.13033 C 2015 AABB V

TRANSFUSION 2015;55;1900–1908 1900 TRANSFUSION Volume 55, August 2015


PLT additive solutions (PASs) can partly replace CPDanticoagulated plasma for storage of PLTs.8 CPD-plasma contains a high concentration of glucose that can be converted into lactate by PLTs, thereby lowering pH, which is associated with reduced recovery after transfusion if pH drops below 6.2.9 Using PASs with acetate instead of glucose prevents this rapid acidification.10-12 Also, PASs can be formulated to modulate specific PLT characteristics; for example, addition of potassium and magnesium to PAS reduces PLT activation.13 An observational study showed that the CCI for PLTs stored in plasma had a mean 1-hour CCI of 10.2 6 6.7 versus 10.0 6 6.3 for PLTs stored in a potassium- and magnesium-containing PAS; a solution without these two ingredients had a mean CCI of 8.6 6 6.4 (p < 0.01).14 Having this in mind, we wondered whether the use of PAS might be able to abrogate the negative effects of PRT on PLT quality, and a four-arm paired study design was developed to allow a good comparison of the various processing, treatment and storage conditions. Buffy coats (BCs) were pooled and split, and care was taken that after preparation, the PLT concentrates were of similar PLT content and volume and that they were stored in the same storage container. SSP1 PAS was used as storage medium in this study; it has acetate as substrate for the PLTs; contains phosphate as a strong buffer to keep pH above 6.2 for good PLT viability; and, further, contains potassium and magnesium, which are known to inhibit PLT activation. While previous studies mainly assessed the use of SSP1 in apheresis PLT concentrates, our study is the first comprehensive paired study comparing plasma versus SSP1 PAS using BC-derived PLT concentrates.

MATERIALS AND METHODS PLT concentrates Whole blood (target volume, 500 mL) was collected in bottom-and-top systems with 70 mL of CPD as anticoagulant and placed under butane-1,4-diol cooling plates until the following morning, when they were centrifuged in a centrifuge (RC-12BP, Sorvall, Kendro, Asheville, NC; 4793 3 g, accumulated centrifugal effect [ACE] 9 3 107) and separated into a RBC unit, a unit of plasma, and a BC of 50 mL (65 mL) with a hematocrit of 42% (66%) using a separator (CompoMat G5, Fresenius, Emmer Compascuum, the Netherlands). Four conditions were compared, that is A) PLT concentrates in plasma, Mirasol treated; B) PLT concentrates in plasma, untreated; C) PLT concentrates in SSP1, Mirasol treated; and D) PLT concentrates in SSP1, untreated. Our current routine PLT concentrates are made from 5 BCs, and a paired four-arm study would require us to pool 20 BCs. However, to compensate for expected PLT loss, as well as to achieve higher PLT concentrations as worst-case condition, we pooled 22 ABO-D-compatible BCs in a 1.3-L container

(F730, Fresenius) using a sterile connection device (TSCDII, Terumo, Tokyo, Japan). To this bag, an additional 1.3-L container was connected, and the content was split equally in half into {A 1 B} and {C 1 D}. To the first split, {A 1 B}, 2 units of plasma (each belonging to one of the 22 original BCs) were added. The content was mixed well, and then equally split into two PLT processing systems (TF*FP0610M1, Terumo), A and B. To the second pool, {C 1 D}, 2 units of SSP1 (300 mL, MacoPharma, Tourcoing, France) were added (aiming at 35% plasma/65% SSP1), and this was also divided over two PLT processing systems, C and D. Air was removed from each system. The BC pools in plasma were centrifuged 4.5 minutes at 1940 3 g (ACE 8.8 3 106) and the units in SSP1 4.25 minutes at 700 3 g (ACE 4.6 3 106) in the centrifuge (RC-12BP, Sorvall). Using the automated separator (CompoMat, Fresenius), the PLTrich supernatant was pushed to the storage container through a leukoreduction filter. From each of the concentrates, a 1-mL sample was taken for PLT counts using a sample pouch. The number of PLTs per concentrate was calculated, and, based on the unit with the lowest PLT count, the number of PLTs was calculated that needed to be removed from each of the other three concentrates. This was then recalculated to a volume of the PLT concentrate that needed to be removed from each of the remaining units. For example, if one unit contained 350 3 109 PLTs, and the other three units contained 400 3 109, and if the PLT concentration was 1 3 109/mL, then from these three PLT concentrates, 50 3 109 would need to be removed, or 50 mL. Volume removal was performed using a sample pouch and a scale; no volume additions were performed. All units conformed to the manufacturer’s requirements for volume and PLT content before the Mirasol treatment. All units were weighed and Control Units B and D were placed on a flat-bed agitator, while in the meantime, Units A and C were Mirasol treated. To each of these units, 35 mL of the riboflavin solution was added and weighed afterward. PLT concentrates in plasma were UV illuminated, with energy delivery calculated per volume (mL) of plasma in each product, according to the manufacturer’s instructions. In all four paired units, a sample site coupler was inserted. To the control units that were not Mirasol treated, based on the volume of riboflavin added to their treated counterpart, an identical volume of saline was added with a syringe. An overview of the various steps is given in Fig. 1. Units were weighed, and after careful mixing, a sample of on average 7 mL was taken with a syringe for in vitro measures. Sampling was repeated on Days 2, 6, and 8.

In vitro measures Volumes were calculated from the net weight divided by the specific gravity (1.027 g/mL for plasma, and 1.014 g/mL for the plasma/SSP1 mixture). pH, blood gases, bicarbonate, Volume 55, August 2015 TRANSFUSION 1901


Fig. 1. Schematic representation of the production of PLT concentrates in the study. Twenty-two BCs were pooled, and divided into A, B and C, D. To A, B, 2 units of plasma were added, and to C, D, 2 units SSP1 were added. These double units were then again divided. The BC pools were then centrifuged, and the PLT-rich supernatant was expressed to a storage container. Based on the lowest PLT count, the volume of the remaining 3 units was reduced to have equal PLT content. To Units A and C, 35 mL of riboflavin solution was added, to Units B and C, 35 mL of saline. Units A and C were illuminated with the appropriate UV dose, and all units were then ready for further investigation.

glucose, and lactate were measured at 37 C with a blood gas analyzer (ABL705, Radiometer, Copenhagen, Denmark). Glucose consumption and lactate production were calculated by dividing the Day 8 minus Day 1 value by the PLT concentration and the number of days. PLT concentration and mean PLT volume were determined with a hematology 1902 TRANSFUSION Volume 55, August 2015

analyzer (XT2000i, Sysmex, TOA, Tokyo, Japan). Residual white blood cells (WBCs) were counted with a counting kit on a flow cytometer (LeukoCount and FACSCalibur, respectively, both from BD Biosciences, Franklin Lakes, NJ). CD62P expression and annexin A5 binding were analyzed by flow cytometry as described elsewhere.12 The hypotonic


shock response (HSR) was performed according to Holme and coworkers15 using an aggregometer (CH 490-4D, Chronolog, Havertown, PA). Plasma was used to dilute samples to prevent PAS-induced differences in the HSR result.16 The PLT responses to ADP (1 mmol/L) and collagen (1 mg/mL) were also determined on the aggregometer in recalcified samples. The mitochondrial potential was determined using JC-1 on the flow cytometer as described before.17 Swirl was judged visually on a scale from 0 (no swirl) to 3 (excellent swirl).

Statistical analysis A repeated-measures analysis of variation was performed using computer software (Instat, Version 3.06, GraphPad, San Diego, CA), followed by Dunnett’s test to compare results during storage with those on Day 1. The various treatments (Mirasol treated or not, plasma or SSP1) were compared using Tukey Kramer’s post test. Differences were considered significant when p values were less than 0.05.

RESULTS The pairing of the units was successful as indicated in Table 1, with almost identical volumes (though significant) and number of PLTs per unit for the various study conditions. Although some volumes were statistically different, this difference of at most 8 mL on a total volume of 360 mL is not likely to impact storage conditions, as the resultant PLT concentrations were 1.03 6 0.06, 1.04 6 0.08, 1.05 6 0.06, and 1.03 6 0.08 3 109 for the four conditions, respectively (not significant). The PLT concentration ranged from 0.89 to 1.21 3 109/mL. The mean 6 SD WBC count was 0.27 6 0.45 3 106 per unit. This rather high mean was caused by one paired experiment with a WBC count of higher than the acceptance criterion of 1 3 106 per unit. However, our guidelines stipulate that 90% of the units must conform (which was the case) in addition to the fact that this low number is not expected to have an effect on PLT storage measures,18 we decided to keep this pair in the analysis. Swirl remained visible, although it was best maintained in the untreated units, which all had a score of 3 on Day 8; Mirasol-treated units in plasma had swirling scores of 1 or 2, averaging 1.3 6 0.5, and those treated in SSP1 had scores of 2 or 3, on average 2.3 6 0.5 on Day 8. Mirasol-treated PLTs both in plasma and in SSP1 showed the largest decline in pH37 C during storage, but all still conformed to national (pH37 C > 6.3)19 and AABB standards (pH > 6.2)20 at the end of storage. The control units in SSP1 showed almost no change in pH. Of note is that the pH of control PLTs in plasma by Day 8 is in the same range as that of Mirasol-treated PLTs in SSP1, an

indication that the use of PAS can diminish some of the effects of PRT. The lowering of pH is caused by the oxidation of glucose into lactic acid, but can be balanced with buffers like bicarbonate and phosphate. The lactate production rate was 0.21 6 0.03 mmol/1011 PLTs/day for Mirasol-treated units in plasma, 0.13 6 0.02 for untreated PLTs in plasma, 0.14 6 0.01 for treated PLTs in SSP1, and 0.09 6 0.01 mmol/1011 PLTs/day for untreated PLTs in SSP1 (all p < 0.001 except control in plasma versus Mirasol-treated in SSP1, p < 0.01). The glucose consumption rates were 0.13 6 0.01, 0.08 6 0.01, 0.09 6 0.01, and 0.05 6 0.01 mmol/1011 PLTs/day for these groups, respectively (all p < 0.001 except control in plasma versus Mirasol-treated in SSP1, not significant). Thus, treated units in plasma have the highest metabolism rate, which is also reflected in the highest decrease in bicarbonate from approximately 17 mmol/L on Day 1 to approximately 4 mmol/L on Day 8. Bicarbonate consumption rates were 0.18 6 0.01, 0.12 6 0.01, 0.05 6 0.01, and 0.01 6 0.01 mmol/1011 PLTs/ day for the four study groups, respectively (all p < 0.001). Despite the production of almost 7 mmol/L lactate, the control units in SSP1 showed almost no depletion of bicarbonate, suggesting either that phosphate has taken up all protons or that the PLTs have formed sufficient bicarbonate from oxidized acetate to neutralize the acidification.11 In the Mirasol-treated group in SSP1, there was 1 unit that had completely exhausted its glucose; all other units still had detectable glucose levels. Mirasol treatment was associated with lower oxygen levels immediately after illumination, and this difference remained detectable throughout storage. However, sufficient oxygen remained present to maintain aerobic metabolism. Carbon dioxide levels were not affected by the Mirasol treatment, but depended on the use of plasma or SSP1.

Functional tests In addition to the biochemical changes, we also studied the functional changes in PLT characteristics. A summary of these tests is shown in Table 2. PLT activation, determined as CD62P expression, was not different immediately after illumination, but increased significantly during storage for Mirasol-treated PLTs both in plasma and in PAS; the control units both in plasma and in SSP1 showed a marginal rise during storage. The apoptotic marker annexin A5 was initially not different among the groups, but was significantly elevated in treated PLTs in plasma on Day 8. Mirasol-treated counterparts in SSP1 had considerably lower values, in fact just slightly above the values found for untreated units in plasma or SSP1. The HSR was immediately lower for Mirasol-treated units in both storage media, but remained constant over time for units in plasma and showed an additional decline in PAS. Untreated units in plasma showed almost no change over a 7-day storage period and some decline for Volume 55, August 2015 TRANSFUSION 1903

1904 TRANSFUSION Volume 55, August 2015

* p < 0.01 versus Day 1. NS = not significant.

Volume, Day 1 (mL) PLTs, Day 1 (3109/U) pH Day 1 Day 6 Day 8 pO2 (mmHg) Day 1 Day 6 Day 8 pCO2 (mmHg) Day 1 Day 6 Day 8 Bicarbonate (mmol/L) Day 1 Day 6 Day 8 Glucose (mmol/L) Day 1 Day 6 Day 8 Lactate (mmol/L) Day 1 Day 6 Day 8 7.04 6 0.02 6.94 6 0.07* 6.90 6 0.06* 90 6 23 59 6 9* 64 6 11* 64 6 1 47 6 5* 45 6 5* 16.6 6 0.7 9.7 6 0.8* 8.3 6 0.5* 17.1 6 0.8 12.8 6 0.9* 11.6 6 0.7* 4.8 6 0.4 12.3 6 .1* 14.1 6 1.7*

7.02 6 0.02 6.75 6 0.05* 6.59 6 0.06* 27 6 7 36 6 9* 45 6 7* 67 6 2 49 6 5* 41 6 4*

16.6 6 0.6 6.4 6 0.5* 3.7 6 0.5*

16.9 6 0.6 10.0 6 0.6* 7.5 6 0.6*

5.0 6 0.4 15.8 6 1.1* 19.8 6 1.9*

B 359 6 14 372 6 34

A 367 6 16 377 6 28


Mirasol treated

PLT concentrates in plasma

3.0 6 0.3 10.3 6 0.7* 13.6 6 1.1*

7.2 6 0.3 2.8 6 0.4* 0.8 6 0.5*

8.2 6 0.2 5.4 6 0.2* 4.5 6 0.4*

31 6 1 29 6 1* 28 6 1*

25 6 8 23 6 9 31 6 10

7.05 6 0.01 6.91 6 0.03* 6.84 6 0.04*

361 6 13 378 6 27


Mirasol treated

3.0 6 0.3 8.1 6 0.7* 9.7 6 1.1*

7.0 6 0.3 4.3 6 0.3* 3.3 6 0.5*

7.9 6 0.3 7.2 6 0.3* 7.3 6 0.5*

29 6 1 26 6 2* 26 6 2*

86 6 20 51 6 10* 59 6 11*

7.06 6 0.01 7.06 6 0.04* 7.08 6 0.05*

364 6 14 373 6 32



PLT concentrates in SSP1

All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 All p < 0.001 except B vs. C, NS

All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 All p < 0.001

All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 except A vs. D, p < 0.01 All p < 0.001

All p < 0.001 except C vs. D, NS All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 except A vs. B, p < 0.05; C vs. D, NS

All p < 0.001 except A vs. C, B vs. D, NS All p < 0.001 except B vs. D, p < 0.01 All p < 0.001 except B vs. D, p < 0.05

All p < 0.001 except B vs. C, p < 0.05; C vs. D, p < 0.01 All p < 0.001 except B vs. C, NS All p < 0.001 except B vs. C, p < 0.01

All NS except A vs. C, B vs. D, p < 0.05; A vs. B, p < 0.001 All NS

Statistics (comparison between Groups A, B, C, and D)

TABLE 1. Storage measures of PLT concentrates in plasma or SSP1, either Mirasol treated or not (mean 6 SD, n 5 12 paired experiments)


* p < 0.01. † p < 0.05 versus Day 1. NS = not significant.

CD62P expression (%) Day 1 Day 6 Day 8 Annexin A5 binding (%) Day 1 Day 6 Day 8 HSR (%) Day 1 Day 6 Day 8 Collagen aggregation (%) Day 1 Day 6 Day 8 ADP aggregation (%) Day 1 Day 6 Day 8 JC-1 (ratio) Day 1 Day 6 Day 8 12 6 4 16 6 2* 17 6 2* 664 19 6 4* 18 6 4* 70 6 15 66 6 6 65 6 5 88 6 10 48 6 31* 62 6 26* 77 6 9 27 6 15* 41 6 25* 12.4 6 1.5 7.7 6 0.7* 8.9 6 2.5*

664 30 6 6* 40 6 5* 32 6 6 37 6 10 36 6 13 91 6 6 45 6 28* 49 6 22* 77 6 11 25 6 13* 30 6 25* 10.3 6 2.3 5.8 6 1.7* 6.1 6 1.3*


A 963 38 6 5* 43 6 4*


Mirasol treated

PLT concentrates in plasma

8.6 6 3.9 4.9 6 2.0† 5.8 6 2.6

29 6 15 3 6 3* 6 6 5*

66 6 21 20 6 19* 30 6 22*

43 6 8 25 6 8* 21 6 8*

664 23 6 5* 23 6 4*

963 37 6 3* 41 6 4*


Mirasol treated

14.5 6 1.7 7.7 6 1.9* 7.9 6 2.7*

24 6 18 1 6 1* 2 6 3*

70 6 22 53 6 28 53 6 28

67 6 11 43 6 7* 51 6 6*

663 16 6 5* 13 6 3*

10 6 4 14 6 4† 14 6 2†



PLT concentrates in SSP1

All NS except C vs. B, D; A vs. D, p < 0.001 All NS except A vs. B, D, p < 0.05; C vs. B,D, p < 0.001 All NS except D vs. A, C, p < 0.05; A vs. B, p < 0.01; B vs. C, p < 0.001

All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 except A vs. B, C vs. D, NS All p < 0.001 except A vs. B, C vs. D, NS

All p < 0.01 except A vs. B, C vs. D, NS; B vs. D, p < 0.05 All NS except C vs. A, B, p < 0.01; C vs. D, p < 0.001 All NS except C vs. D, p < 0.05; B vs. C, p < 0.001

All p < 0.001 except B vs. D, NS; A vs. C, p < 0.05 All p < 0.001 except A vs. D, NS; A vs. C, p < 0.01 All p < 0.001

NS All p < 0.001 except B vs. D, NS; B vs. C, p < 0.05 All p < 0.001 except B vs. D, p < 0.01

All NS except D vs. B, C, p < 0.01; B vs. A, C, p < 0.001 All p < 0.001 except A vs. C, B vs. D, NS All p < 0.001 except A vs. C; B vs. D, p < 0.05

Statistics (comparison between Groups A, B, C, and D)

TABLE 2. Functional measures of PLT concentrates in plasma or SSP1, either Mirasol treated or not (mean 6 SD, n 5 12 paired experiments)


Volume 55, August 2015 TRANSFUSION 1905


untreated PLTs in PAS. For the aggregation assay, we used collagen and ADP as physiologic stimulants. Upon collagen stimulation, PLTs in PAS, both treated and untreated, showed significantly lower values immediately after preparation. An additional decline is seen for all study conditions, but was most pronounced for the treated units in SSP1. Upon ADP stimulation, a similar observation was made, but responsiveness to ADP almost disappeared for units in PAS when stored. Finally, JC-1 was studied to determine the mitochondrial potential, where a lower ratio indicates a lower mitochondrial potential. Mirasol treatment affected the JC-1 signal to some extent, but also storage time was of influence on the JC-1 signal; changes in JC-1 were not different between treated units in plasma or in PAS (A vs. C and B vs. D). The one unit with a glucose of 0 (in the Mirasoltreated group in SSP1) had a JC-1 signal of 6.0, which is a little above the average of the entire group.

DISCUSSION This study investigated whether the use of PAS could abrogate the effects of Mirasol PRT on BC-derived PLT concentrates. Our data suggest that SSP1 lowers glucose metabolism of treated units to levels of untreated units in plasma. Less lactate is produced, and also the pH of Mirasol-treated PLTs in SSP1 is maintained at a level similar to untreated PLTs in plasma. Annexin A5 showed only a small difference between treated PLTs in SSP1 and untreated PLTs in plasma, all suggesting that, indeed, the use of a PAS can prevent some of the damage induced by Mirasol PRT. However, PAS could not abolish all effects of PRT, such as PLT activation after 8 days of storage, which was almost identical for treated PLTs in plasma or SSP1. The HSR score of the PLTs was lower after 8 days when SSP1 was used rather than plasma; also the responses to collagen and ADP were lower, but this was mainly due to the use of PAS itself, not by the PRT process. Previous studies have shown that Mirasol-treated PLTs have elevated glycolytic rates.21 A subsequent study revealed that this was not due to loss of mitochondrial integrity, as the JC-1 signal was not changed,22 and mitochondrial activity was not affected by the PRT procedure.23 Our study confirms that PRT does not affect the JC-1 signal for PLTs stored in plasma, although for PAS there does seem to be an effect. Storage of the PLT concentrates gives additional changes in the JC-1 signal. Li and colleagues23 also found that the ATP concentration was lower in treated than in untreated units. This altogether suggests that the PLTs have functioning mitochondria but that there is an elevated demand for ATP, which then induces the need to oxidize more glucose into lactate. Picker and colleagues24,25 confirmed these findings in their studies evaluating the effect of Mirasol treatment on apheresis PLTs in plasma and in SSP1. For PLTs in 1906 TRANSFUSION Volume 55, August 2015

plasma,24 they found lower HSR response, lower swirling score, poorer aggregation upon stimulation with ristocetin, higher CD62P expression, and more depolarized PLTs when Mirasol treated. However, they had different PLT concentrations between the groups, namely, 1.07 6 0.11 3 109/mL for the Mirasol-treated units and 1.24 6 0.13 3 109/mL for control PLTs, which might favor storage variables in the Mirasol group. A subsequent study with apheresis PLTs in SSP1 (35% residual plasma)25 showed the same decline in in vitro quality as PLTs in plasma; in the design of that study they used PLT concentrates with a similar PLT concentration of 0.99 6 0.08 3 109/mL in both groups which makes comparison easier. HSR was lower, swirl was lower, and CD62P expression was considerably higher as was annexin A5 binding, and the percentage of depolarized PLTs was higher when treated. Most notable is the fact that the absolute values of control PLTs in plasma and the Mirasol-treated PLTs in SSP1 were almost similar; for example, HSR was 76 6 4% for control units in plasma24 and 74 6 16% for Mirasol-treated units in SSP1,25 but the difference in PLT concentration could not be ruled out as explanation. Our study indeed shows that HSR was lower for Mirasol-treated PLTs in SSP1. A similar observation can be made for CD62P expression (46 6 9% for untreated PLTs stored for 7 days in plasma24 vs. 57 6 11% for Mirasol-treated PLTs stored for 7 days in SSP125). These data suggested that SSP1 might prevent part of the negative effect of PRT on in vitro measures, and again our study indicates that there are indeed considerable differences between Mirasol-treated PLTs in SSP1 versus untreated ones in plasma. In this strictly controlled study with minimal variation in volume and PLT concentration, we find that the lower glucose concentration and the replacement of glucose for acetate as fuel for the PLTs reduces lactate production from on average 0.13 (untreated in plasma) to 0.09 mmol/1011 PLTs/day (untreated in SSP1). PRT then induces an increase in metabolism to 0.14 mmol/1011 PLTs/day for treated PLTs in SSP1. This is reflected in similar pH for treated units in SSP1 and untreated units in plasma. Therefore, we postulate that the increase in metabolism induced by the Mirasol procedure can be counteracted by the use of PAS, more specifically the replacement of glucose for acetate as PLT fuel. Consequently, the better storage variables are reflected as a more stable pH and no effect on annexin A5. In addition, a comparison of two PASs showed that phosphate helps maintain the pH in Mirasol-treated PLT concentrates.26 In contrast, the elevated CD62P expression seems to be initiated by a different pathway that is more directly targeted by the Mirasol procedure. Despite the fact that a rise in CD62P expression and annexin A5 binding are almost always seen together, in our study we find a decoupling of the two processes, which has been described recently in vitro using different stimuli.27 As reviewed in detail


elsewhere,28 mitochondrial depolarization and annexin A5 positivity are linked through an apoptosis pathway: apoptosis is triggered by, for example PRT; this lowers the membrane potential which in turn activates caspases, which ultimately lead to phosphatidylserine exposure, which is measured by annexin A5 binding. Thus, PLT metabolism and apoptosis are not always linked to degranulation and activation, although they might be synergistic or happen together when other pathways are performing at full power. Our study shows that subtle control of metabolism may help maintain PLT survival by keeping annexin A5 low, while other effects (basal degranulation) are independent. How the use of PAS inhibits PLT apoptosis is an area of further research. A decline in oxygen was seen immediately after Mirasol treatment. This might suggest that this is due to riboflavininduced photooxidation, but the Mirasol procedure in fact uses an oxygen-independent electron transfer,29 so this is not a likely explanation. Most probably, the lower oxygen levels are caused by escape of oxygen during multiple transfers from container to gas-permeable container. HSR was affected by the PRT immediately after the procedure and remained to show results less than controls throughout storage. The HSR is not a real functional test, but has been shown to have a good correlation with in vivo recovery and survival.30 The response to collagen stimulation was lower for Mirasol-treated PLTs than for control units. The use of PAS also seemed to affect the collagen response, and so a cumulative negative effect is seen for PRT and for PAS. This was also the case for the response after stimulation with ADP. The response in SSP1 is lower due to the absence of apyrase (which is normally present in plasma), causing desensitization of PLTs.31 PRT enhances this effect. We found that the JC-1 signal was lower for Mirasol-treated PLTs versus their untreated counterparts, indicating mitochondrial depolarization of the PLTs. Finally, swirl was better in treated units in SSP1 than those in plasma, although still not as good as in untreated PLTs stored in either plasma or PAS. Previous studies, also from our own institution32,33 have shown that PLTs that underwent PRT are more activated than their controls. Whether or not this is an unwanted phenomenon is a matter of debate, as activated PLTs are possibly more efficient if a patient has vascular damage. The relationship between in vitro measures and recovery and survival is weak.34 There are significant correlations for a number of in vitro measures, but the r2 is never higher than 0.5; annexin A5 probably has the best correlations with an r2 value of 0.26 for PLT recovery and r2 value of 0.41 for PLT survival. Considering the ability of PAS to moderate the increased annexin A5 binding in Mirasol-treated PLTs, we can only speculate that the use of PAS can further improve recovery and survival. There are even scarcer data on the relation between in vitro

measures and CCI, let alone bleeding tendencies. It is known that the CCI for Mirasol-treated PLTs in plasma is lower than untreated controls,7 but if and to what extent this affects bleeding tendency in patients is not known. In any case, our center, in collaboration with hospitals in Norway and Canada, is conducting a randomized clinical trial studying the effect of Mirasol treatment on the ability of PLTs to stop or prevent bleeding in patients, the PREPAReS trial. In this trial, samples are taken from the PLT concentrates before transfusion for in vitro analysis, enabling us to establish the relation between laboratory tests and CCIs or bleeding risk. Finally, the question remains whether the use of plasma or PAS affects the degree of pathogen inactivation. Both for bacteria35 and for viruses,36 no meaningful difference could be shown, so there is no significant bias in pathogen reduction performance induced by the storage medium of the PLTs. In summary, the use of PAS for production and storage of BC-derived PLT concentrates that are subsequently Mirasol treated have some beneficial action as pH is corrected because lactate production is lowered, and annexin A5 is preserved, but the PLT activation is as high as in treated PLTs in plasma, HSR is diminished, and the response to collagen is low. Further clinical data need to be gathered to prove if the use of PAS is beneficial to use for PLTs in combination with Mirasol PRT. ACKNOWLEDGMENTS We thank Terumo BCT for supplying illuminators and disposables at no cost. We thank Dr Laura Gutierrez for critical revision of the manuscript and for helpful suggestions.

CONFLICT OF INTEREST The authors have disclosed no conflicts of interest.

REFERENCES 1. Dwyre DM, Fernando LP, Holl PV. Hepatitis B, hepatitis C and HIV transfusion-transmitted infections in the 21st century. Vox Sang 2011;100:92-8. 2. Alter HJ. Pathogen reduction: a precautionary principle paradigm. Transfus Med Rev 2008;22:97-102. 3. Fast LD, DiLeone G, Marschner S. Inactivation of human white blood cells in platelet products after pathogen reduction technology treatment in comparison to gamma irradiation. Transfusion 2011;51:1397-404. 4. Asano H, Lee CY, Fox-Talbot K, et al. Treatment with riboflavin and ultraviolet light prevents alloimmunization to platelet transfusions and cardiac transplants. Transplantation 2007;84:1174-82. 5. Goodrich RP, Gilmour D, Hovenga N, et al. A laboratory comparison of pathogen reduction technology treatment and Volume 55, August 2015 TRANSFUSION 1907


culture of platelet products for addressing bacterial contamination concerns. Transfusion 2009;49:1205-16. 6. AuBuchon JP, Herschel L, Roger J, et al. Efficacy of apheresis platelets treated with riboflavin and ultraviolet light for pathogen reduction. Transfusion 2005;45:1335-41. 7. Mirasol Clinical Evaluation Study Group. A randomized controlled clinical trial evaluating the performance and safety of platelets treated with MIRASOL pathogen reduction technology. Transfusion 2010;50:2362-75. 8. Azuma H, Hirayama J, Akino M, et al. Platelet additive solution—electrolytes. Transfus Apher Sci 2011;44:277-81. 9. Murphy S, Sayar SN, Gardner FH. Storage of platelet concentrates at 22 degrees C. Blood 1970;35:549-57. 10. Fijnheer R, Veldman HA, van den Eertwegh AJ, et al. In vitro evaluation of buffy-coat-derived platelet concentrates stored in a synthetic medium. Vox Sang 1991;60:16-22. 11. Bertolini F, Murphy S, Rebulla P, et al. Role of acetate during platelet storage in a synthetic medium. Transfusion 1992;32: 152-6. 12. van der Meer PF, Kerkhoffs JL, Curvers J, et al. In vitro comparison of platelet storage in plasma and in four platelet additive solutions, and the effect of pathogen reduction: a proposal for an in vitro rating system. Vox Sang 2010;98: 517-24. 13. de Wildt-Eggen J, Schrijver JG, Bins M, et al. Storage of platelets in additive solutions: effects of magnesium and/or potassium. Transfusion 2002;42:76-80. 14. Tardivel R, Vasse J, Gaucheron S, et al. A comparative study of the efficiency of plasma and additive solution preserved platelet concentrates [abstract]. Vox Sang 2012; 103(Suppl 1):247. 15. Holme S, Moroff G, Murphy S. A multi-laboratory evaluation of in vitro platelet assays: the tests for extent of shape change and response to hypotonic shock. Biomedical Excellence for Safer Transfusion Working Party of the International Society of Blood Transfusion. Transfusion 1998;38: 31-40. 16. VandenBroeke T, Dumont LJ, Hunter S, et al. Platelet storage

21. Ruane PH, Edrich R, Gampp D, et al. Photochemical inactivation of selected viruses and bacteria in platelet concentrates using riboflavin and light. Transfusion 2004;44:877-85. 22. Li J, de Korte D, Woolum MD, et al. Pathogen reduction of buffy coat platelet concentrates using riboflavin and light: comparisons with pathogen-reduction technology-treated apheresis platelet products. Vox Sang 2004;87:82-90. 23. Li J, Lockerbie O, de Korte D, et al. Evaluation of platelet mitochondria integrity after treatment with Mirasol pathogen reduction technology. Transfusion 2005;45:920-6. 24. Picker SM, Steisel A, Gathof BS. Effects of Mirasol PRT treatment on storage lesion development in plasma-stored apheresis-derived platelets compared to untreated and irradiated units. Transfusion 2008;48:1685-92. 25. Picker SM, Tauszig ME, Gathof BS. Cell quality of apheresisderived platelets treated with riboflavin-ultraviolet light after resuspension in platelet additive solution. Transfusion 2012; 52:510-6. 26. Cookson P, Thomas S, Marschner S, et al. In vitro quality of single-donor platelets treated with riboflavin and ultraviolet light and stored in platelet storage medium for up to 8 days. Transfusion 2012;52:983-94. 27. Gyulkhandanyan AV, Mutlu A, Freedman J, et al. Selective triggering of platelet apoptosis, platelet activation or both. Br J Haematol 2013;161:245-54. 28. Leytin V. Apoptosis in the anucleate platelet. Blood Rev 2012; 26:51-63. 29. Cadet J, Decarroz C, Wang SY, et al. Mechanisms and products of photosensitized degradation of nucleic acids and related model compounds. Isr J Chem 1983;23:420-9. 30. Holme S, Moroff G, Whitley P, et al. Properties of platelet concentrates prepared after extended whole blood holding time. Transfusion 1989;29:689-92. 31. Keuren JF, Cauwenberghs S, Heeremans J, et al. Platelet ADP response deteriorates in synthetic storage media. Transfusion 2006;46:204-12. 32. Middelburg RA, Roest M, Ham J, et al. Flow cytometric assessment of agonist-induced P-selectin expression as a

solution affects on the accuracy of laboratory tests for platelet function: a multi-laboratory study. Biomedical Excellence

measure of platelet quality in stored platelet concentrates. Transfusion 2013;53:1780-7.

for Safer Transfusion Working Party of the International Soci-

33. Zeddies S, De Cuyper IM, van der Meer PF, et al. Pathogen

ety of Blood Transfusion. Vox Sang 2004;86:183-8. 17. Verhoeven AJ, Verhaar R, Gouwerok EG, et al. The mitochon-

reduction treatment using riboflavin and ultraviolet light impairs platelet reactivity toward specific agonists in vitro.

drial membrane potential in human platelets: a sensitive parameter for platelet quality. Transfusion 2005;45:82-9. 18. Pietersz RN, de Korte D, Reesink HW, et al. Preparation of leukocyte-poor platelet concentrates from buffy coats. III. Effect of leukocyte contamination on storage conditions. Vox Sang 1988;55:14-20. 19. De Korte D, editor. Richtlijn bloedproducten [Guideline blood products]. 6th ed. Amsterdam: Sanquin Blood Supply; 2013. 20. Fung MK, Grossman BJ, Hillyer C, Westhoff CM, editors. Technical manual. 18th ed. Bethesda (MD): American Association of Blood Banks; 2014. 1908 TRANSFUSION Volume 55, August 2015

Transfusion 2014;54:2292-300. 34. Slichter SJ, Corson J, Jones MK, et al. Exploratory studies of extended storage of apheresis platelets in a platelet additive solution (PAS). Blood 2014;123:271-80. 35. Keil SD, Hansen E, Gilmour DI, et al. Treatment of platelet products in additive solution with riboflavin and UV light: effectiveness against bacterial contamination [abstract]. Transfusion 2009;39(Suppl):227A. 36. Keil SD, Bengrine A, Bowen R, et al. Inactivation of viruses in platelet and plasma products using a riboflavin-and-UV-based photochemical treatment. Transfusion 2015;55:000-00.

Riboflavin and UV light treatment of platelets: a protective effect of platelet additive solution?

Pathogen reduction technologies (PRTs) increase the safety of the blood supply, but are also associated with cell damage. Our aim was to investigate t...
186KB Sizes 0 Downloads 11 Views