Platelets: Testing, Dosing and the Storage Lesion—Recent Advances

The demand for platelet transfusions continues to grow. Several complementary approaches may help meet this demand in the future. First, platelet bacterial testing is beginning to allow the extension of platelet storage beyond 5 days. Studies are also underway aimed at better preserving viability and function during ex vivo platelet storage: additive solutions and other approaches are being developed to try to negate the “platelet storage lesion.” Finally, new approaches to dosing platelets may help extend the limited supply.

Over the years, improvements in donor screening nearly eliminated hepatitis B virus, hepatitis C virus and HIV from the US blood supply.^1,2 As the transfusion-transmission rates of these agents dropped, platelet bacterial contamination assumed a new prominence as the most frequent infectious risk of transfusion. Unlike red cells, which are stored under refrigeration, platelets are best stored at room temperature. If platelets are refrigerated prior to being transfused, they are rapidly cleared from the recipient’s circulation.³ While room temperature storage allows transfused platelets to circulate in vivo, it has the downside of promoting bacterial growth. Because of this risk, platelet storage is ordinarily limited to only 5 days, making platelet inventory management extremely challenging.

In the 1990s, numerous studies demonstrated that contaminating bacteria, usually representing skin flora from the donor, could be cultured out of approximately 1/3000 platelet units.⁴ Clinically apparent septic transfusion reactions were thought to occur following ~1/25,000 platelet transfusions, although there is considerable uncertainty around this point estimate.⁴ In 2002, an open letter calling for immediate action to reduce the risk of platelet bacterial contamination⁵ led to the proposal of AABB standard 5.1.5.1. The standard states:

All AABB accredited blood banks were required to meet standard 5.1.5.1 by March 1, 2004. How this standard is being met varies by facility. As of now, three culture-based bacterial detection systems are licensed in the US for quality control use: BacT/Alert (BioMerieux), eBDS (Pall) and Scansystem (Hemosystem). “Quality control” in this context means that the test is licensed to verify that the platelet collection process does not introduce contaminating bacteria more frequently than a predefined rate. These tests are not yet licensed as “release tests,” which by definition are used to confirm that a platelet product being issued is not contaminated with bacteria. The BacT/ALERT system, used by most centers performing culture-based screening, works by continuously monitoring for bacterial within culture bottles. Platelet products production of CO₂ are generally sampled on day 1 after collection. The samples are cultured for a period of time, typically 24 hours, and if the cultures fail to produce abnormal levels of CO₂, the product is released into inventory. Both aerobic and anaerobic cultures may be performed, although it is known that aerobic organisms cause the vast majority of septic reactions. The Pall eBDS (enhanced bacterial detection system) detects oxygen consumption by aerobic bacteria. An in the air oxygen sensor is used to measure the residual O₂ above a platelet sample 24 hours after sampling. The Hemosystem Scansystem also requires a culturing step, but this device detects bacteria directly by using a solid-phase fluorescent cytometric method. In the US, these systems are only used to test single donor apheresis platelet units. By current US Food and Drug Administration (FDA) rule, whole blood–derived platelets must be pooled prior to issue as opposed to prestorage.^7,–9 Individually culturing multiple concentrates comprising a single platelet dose would be prohibitively cumbersome and expensive, so for whole blood–derived platelets, non-licensed, non-culture methods such as pH and glucose testing (using a dipstick or automated analyzer) and Gram staining are performed. However, the sensitivity and specificity of these non-culture methods is inferior to culture-based methods.^10,11 Prestorage pooling of whole blood–derived platelet concentrates is routinely done in a number of countries. Studies now in progress, combined with data from other countries, will likely lead to licensing of prestorage pooled whole blood–derived platelets in the US, thus allowing an equivalent level of surveillance for all platelet products.

Standard 5.1.5.1 has been in effect long enough to begin to assess its impact on platelet transfusion safety. Table 1  summarizes three large-scale surveillance studies of septic reactions that were performed prior to the implementation of standard 5.1.5.1: a Johns Hopkins University study,¹² the French BACTHEM study,¹³ and the US BaCon study.¹⁴ In the Hopkins and BACTHEM studies, the risk of a clinically apparent septic transfusion reaction ranged from 1/2500–1/31,000 platelet transfusions. Fatalities were observed following ~1/5–1/10 septic reactions. Bacterial contamination appeared to occur less often with single donor platelets than with pooled random donor platelets.^12,13 The prevalence of septic reactions was dramatically lower in the BaCon study than in the other two studies; it has been argued that the BaCon study was limited by an underestimation bias.⁴ 

A large post-standard 5.1.5.1 study is also summarized in Table 1 .¹⁵ Over a 10-month period, apheresis platelets collected at 36 American Red Cross blood centers were cultured 24 hours after collection using the BacT/ALERT system. Products were eligible for release after 12 hours without a positive culture. In all, 350,658 units were tested. There were 226 initially positive cultures (1/1552), of which 68 were confirmed as true positive (1/5157). During the study period, 3 platelet transfusions, screened as negative by BacT/ALERT, resulted in “high probability” septic reactions (1 fatal) due to coagulase-negative Staphylococcus. In contrast, over an identical period of time prior to bacterial screening, there were 12 high probability septic reactions (2 fatal). A similar rate of contaminated products was seen in a study by Holme and colleagues, who identified 23 confirmed positives out of 118,067 platelet units tested using the Pall eBDS system (1/5133).¹⁶ Similar contamination rates have also been reported by European centers using the BacT/ALERT system.¹⁷ Overall, culture-based bacterial screening appears to reduce, but not eliminate, the risk of septic reactions. Reducing the risk even further may ultimately require alternative approaches, such as pathogen reduction.^15,18,–20 

While the current culture systems have increased platelet safety, there are a number of costs: higher prices for platelets, 1–2% product loss from sampling and 12–30 hour delays in platelet product release, potentially affecting platelet availability.²¹ It is hoped that these costs will be offset by the benefits of being able to extend platelet storage beyond the 5-day limit. The FDA has established two criteria by which the shelf life of platelets may be extended up to 7 days. First, products must be stored in containers approved for 7-day storage (these are available). Second, products must test negative for bacteria with an FDA-approved release test. The FDA has issued specific guidelines under which the current bacterial detection devices could be validated for release testing.²² In post-marketing surveillance studies, cultures will need to be performed following platelet outdating (day 7) to confirm the initial (day 1) negative culture. Approval of a culture-based release test will require demonstrating that a negative day 1 culture will only give rise to a true positive day 7 culture in 1/10,000 cases, with 95% confidence that the actual risk is < 1/5000. (A risk of < 1/5000 would be considered to be lower than the current estimated risk for 5-day platelets.) To achieve this high degree of statistical confidence, a very large study size will be required, approximately 50,000 units. Gambro BCT has initiated such a study, the “Post Approval Surveillance Study of Platelet Outcomes, Release Tested” (PASSPORT) study. In this protocol, apheresis platelets collected using the Spectra or Trima devices are cultured 24 hours after collection using the BacT/ALERT system. Both aerobic and anaerobic cultures are performed. After 24 hours, the products are made available for clinical use with a 7-day shelf life. Any products that outdate are re-cultured on day 8. Fifty thousand outdated products are to be tested.²¹ 

Because of their high sensitivity (detection limit of 1–10 CFU/mL), culture-based systems are the current gold standard for platelet bacterial detection. However, culturing platelet units is both cumbersome and expensive. Also, platelet cultures use up 4–8 mL of platelet product that would otherwise have gone to the patient, and the 24-hour incubation period shortens an already brief shelf life. Ideally, a sensitive, inexpensive test for bacteria could be performed rapidly on a small volume of platelet product just before issue. Several such tests are under development, although none is licensed as of this writing.^23,24 Current-generation rapid bacterial tests have a turnaround time of a few hours or less, and use < 1 mL of platelet product. They have a lower sensitivity than culture, however—on the order of 10³–10⁴ CFU/mL. In principle, rapid bacterial tests could be used for quality control, as an adjunct to another test e.g., initial culture, or as a stand-alone release test. Licensing criteria for rapid bacterial tests are being developed by the FDA.²⁵ 

If platelet storage is routinely extended past 5 days, it will be important to understand how well 7-day-old platelets, (or perhaps even older platelets) will function clinically. It has long been recognized that platelet viability and function decline over time, the so-called “platelet storage lesion.”^26,27 The most compelling data pointing to damage incurred during platelet storage come from radiolabeling studies in autologous volunteer platelet donors. Most investigations have shown that platelets transfused following storage have a significantly decreased recovery and survival relative to fresh platelets.^26,28,29 For example, day 5 and day 7 platelets have been compared directly in radio-labeling studies. Day 5 platelets were found to have a mean recovery of 63% and a mean survival of 160.8 hours, compared with day 7 platelets, which had a recovery and survival of 53.9% and 133.6 hours, respectively.³⁰ 

Numerous morphologic, biochemical and functional derangements occur during platelet storage. Platelets continue to be metabolically active at room temperature. Products of metabolism such as lactate accumulate, and the pH falls. It has been shown that if the pH drops below 6.0–6.2, survival in vivo is severely diminished.²⁶ Platelets also tend to become activated during storage. Over time, an increasing fraction of platelets in a concentrate will change from a discoid (resting) shape to spherical. Mediators of thrombosis such as β-thromboglobulin and platelet factor 4 accumulate in the storage medium, reflecting granule release. Platelet surface levels of P-selectin (CD62P), another platelet activation marker, also increase during storage. Finally, functional derangements are observed in vitro. Platelet aggregation responses to a number of agonists drop significantly during storage. Still, beyond a few extreme cases (e.g., pH < 6.0) the significance of the in vitro abnormalities observed following platelet storage remains unclear. First, no in vitro test has yet been validated to reliably predict platelet survival in vivo.³¹ Also, there is good evidence suggesting that many of the platelet storage abnormalities seen in vitro are actually reversible upon transfusion.³² 

In practice, in vitro tests are used to initially assess platelet preparations, but licensing of any new platelet product currently requires in vivo survival studies in autologous volunteers. The FDA now requires that new candidate platelet products must be directly compared to fresh whole blood–derived platelets in dual radiolabeling experiments. To be licensed, a candidate platelet product will need to demonstrate at least 66% of both the recovery and survival of fresh platelets.

In the US, platelets are currently stored in plasma only. Optimized synthetic storage media might help attenuate the platelet storage lesion, thereby facilitating extended storage. Numerous platelet additive solutions have been formulated with this idea in mind. Because additive solutions replace 70–80% of the plasma in the original platelet unit, these formulations are predicted to have additional benefits: reduced allergic and febrile transfusion reactions,³³ decreased transfusion of unwanted antibodies (e.g., ABO, HLA) and increased plasma made available for fractionation.³⁴ Additive solutions could afford technical benefits for pathogen reduction methods as well (PAS-III, or “Inactisol” was created for the EuroSPRITE pathogen reduction study).

The composition of several platelet additive solutions is shown in Table 2 . The latest-generation solutions are currently licensed in Europe; these include PAS-IIIM (MacroPharma) and Composol (Fresenius). PlasmaLyte A (Baxter), initially formulated as an intravenous replacement fluid, is the only similar product that is licensed in the US. In an additive solution unit, the final medium contains 20–30% donor plasma. This carried-over plasma provides glucose for platelet metabolism. PAS-IIIM, Composol and PlasmaLyte A also contain acetate, which serves as a second metabolic fuel. Acetate has the added benefit of acting as a buffer. Bicarbonate, the most important physiologic buffer, has been found unsuitable for product manufacturing and so is no longer included in additive solutions. Magnesium and potassium are present in PAS-IIIM, Composol and PlasmaLyte A. These electrolytes inhibit platelet activation and aggregation, although how they work is unclear.³⁴ Van der Meer and colleagues³⁵ compared the in vitro storage characteristics of pooled buffy coat platelets stored for up to 12 days in 100% plasma, or in mixtures of plasma with PAS-II, PAS-III, PAS-IIIM and Composol. They observed that several in vitro markers of platelet quality (pH ≥ 6.8, glucose consumption, lactate production) were reasonably well preserved for 9–12 days in platelets stored either in 100% plasma, or in PAS-IIIM (30% plasma) or Composol (35% plasma). In general, those additive solution preparations containing higher concentrations of plasma performed better, probably due to the buffering capacity of plasma. As noted above, the value of in vitro platelet quality markers is limited. Thus, in vivo survival studies of additive solution–stored platelets are now being undertaken.

Rather than modifying the storage medium to improve and extend platelet storage, another possibility is to modify the platelets themselves. One such strategy, proposed by Hoffmeister and colleagues, is to treat platelets so that they can be stored under refrigeration without a subsequent loss of viability.³⁶ Platelet refrigeration would significantly reduce the bacterial growth risk, potentially allowing for extended storage. Also, refrigerated platelets could theoretically have functional advantages over room temperature platelets, although this is unproven in vivo.³⁷ When murine platelets are cooled to 4°C for 2 hours, von Willebrand factor receptor (GPIb) complexes cluster together irreversibly on the platelet surface. In this clustered conformation, platelet GPIb molecules are specifically bound by α_Mβ₂ integrin receptors on liver macrophages. In turn, the liver macrophages phagocytize the platelets, pulling them out of the circulation.³⁸ The critical binding interaction involves an α_Mβ₂ integrin receptor lectin domain that recognizes β-N-acetylglucosamine (β-GlcNAc) residues present on GPIb N-linked oligosaccharides. In mice, it has been possible to inhibit this binding—and thus prevent the clearance of chilled platelets—by coating platelet β-GlcNAc residues with galactose.³⁶ Whether this strategy can be translated into human platelet products suitable for refrigerated storage remains to be seen.

Apart from extending platelet storage, there exist other strategies to stretch the available supply of platelets. One very successful maneuver was simply lowering the transfusion trigger for prophylactic platelet transfusion from 20,000/μL to 10,000/μL. This change, found to be safe in a number of studies,^39,40 is thought to have reduced the number of platelet doses transfused in the US by 20–30%,⁴¹ although no large-scale studies have been performed to determine the actual impact of the altered transfusion threshold.

A complementary approach to lowering the trigger is altering the dose of prophylactic platelets provided. There are thought to be two separate clearance mechanisms for platelets. Most platelets undergo senescence following a lifespan within the circulation of ~10 days. There is also evidence for a fixed daily loss of platelets that occurs independent of platelet age. Platelets exiting the circulation via this second route (estimated to be 7.1 × 10⁹ platelets/L/day) are postulated to function in maintaining vascular integrity.⁴² In principle, a low dose of platelets could be used to fulfill this daily requirement, and mathematical modeling suggests that using a low dose (e.g., pools of 3 whole blood–derived platelets versus the current 6) would save ~22% of prophylactic platelets over time.⁴³ On the other hand, higher prophylactic platelet doses would be predicted to reduce the total number of transfusion events.⁴⁴ 

To determine empirically the optimal prophylactic platelet dose, a randomized controlled trial is now being conducted by the Transfusion Medicine/Hemostasis Clinical Trials Network.⁴⁵ In this study, termed “PLADO” (platelet dosing), thrombocytopenic patients are randomized to one of three arms: medium (standard) platelet dose (2.2 × 10¹¹ platelets/m²); low dose (1.1 × 10¹¹ platelets/m²); or high dose (4.4 × 10¹¹ platelets/m²). Patients receive prophylactic transfusions for a morning platelet count of ≤ 10,000/mL. The primary end point is the percentage of patients with Grade 2 or higher bleeding in each arm. The plan is to enroll a total of 1350 patients. This study will be completed in 2007 and is expected to have a broad impact on how prophylactic platelets are administered moving forward.