92nd Meeting, September 10-11, 2008

Rockville, MD


Issue Summary


Topic I. Strategies to Enhance Bacterial Safety of 7 day Platelets for Transfusion


Issue:FDA seeks the advice of the Committee on the development of new testing paradigms that will reduce the rates of bacterial contamination and septic transfusion associated with platelet products




Platelets are currently stored at room temperature up to 5 days.They can be a medium for bacterial growth if the platelet product is contaminated during collection.Bacteria introduced in small numbers into a platelet bag can proliferate to bacterial concentrations of 1012 CFU/mL and greater.While some contaminated units may display visible changes, often they do not appear much different from uncontaminated units.When transfused, contaminated units can have grave results.


FDA extended platelet shelf life to 7 days in the mid 1980ís.However, after approximately 1.5 years, FDA, on advice from BPAC, reduced the shelf life back to 5 days because of increased septic transfusion reaction rates from the 7 day platelets.Since then several advancements made it possible to re-introduce 7 day platelets on the market.Devices to detect bacterial contamination of platelets were developed and cleared by the FDA for quality control of platelet products.The device manufacturers did not seek a claim for a release test and thus no clinical study was done to validate such a claim.However, the use of culture based devices has become routine for quality control screening of 5 day platelets and a large body of data became available on detection of contaminated products when these were sampled 24 hours after collection.The contamination rate was found to be about 1/5,000, which was generally thought to be the contamination rate based on several small studies that screened collected platelet products. It was also thought that the use of these devices early in the storage period would be sufficient to reduce the contamination rate of 7 day platelets.


Several companies validated their collection containers for platelet storage up to 7 days.One company, Gambro, collected performance data on the use of BacT Alert for quality control testing of platelets and prepared a study proposal for establishing uniform use of the bacterial detection device to screen platelets.Gambro received clearance from the FDA for marketing 7 day apheresis platelets with labeling that specified the use of the BacT Alert culture system on a platelet sample obtained 24 hours after the collection.Additionally, FDA required that the users participate in a post market study to validate the performance of the BacT Alert device for reducing the contamination rate of 7 day platelets.This study became known as the PASSPORT (Post Approval Surveillance Study of Platelet Outcomes, Release Testing) study and was eventually joined by a second sponsor, Fenwal Corporation.


PASSPORT Study Design


The objective of the study was to demonstrate that the point estimate for the contamination rate of 7 day apheresis platelets is 1/10,000 or less, with the upper 95% confidence limit not greater than 1/5,000.†† The 1/5,000 rate was chosen because this was thought to be the contamination rate of untested 5 day products.The study had a standardized testing protocol that used two culture bottles (aerobic and anaerobic), each inoculated with 4-5 ml of sample collected 24 to 36 hours after venipuncture.There was no requirement for diversion of initially collected blood although many collection centers used this practice anyway.The products were held for 24 hours after testing and, if negative at that time, were released for transfusion.The inoculated culture bottles were monitored in the incubator for 7 days.The products were transfused for up to 7 days.Outdated products were sent back to the collection center for retesting (surveillance).The study was planned for testing of 50,000 outdated units and allowed less than 5 true positives in the outdated units (5/50,000 = 1/10,000).


Over the course of approximately 2.5 years the study tested 320,983 products at 24 hours and 4,369 products at outdate.The detected contamination rate was approximately 1/5,000 at 24 hours and 1/1,400 at outdate (3/4,369). These results showed that testing of 24 hour platelets had a high false negative (FN) rate (approx. 74%)


††††††††††††††††††††††† ††††††††††† ††††††††††† Day 1††††††††††††††††††††††††† Day 7

Contamination rate†††††††††††††††† 1/4,273††††††††††††††††††††††† 1/1,456

††††††††† N=†††††††††††††††††††††††††††††† 320,983†††††††††††††††††††††† 4,369


Approximately two years into the study after the first interim analysis the sponsors became aware of the unexpectedly high contamination rate of outdated products.They became concerned over the safety of 7 day platelets and decided to stop production of 7 day platelets even though the study tested only a fraction of the planned 50,000 outdated units.Two main conclusions emerged from the first interim analysis of the data: 1) The number of bacterially contaminated units identified at outdate (2 out of 2,571 tested) made it very unlikely that the study would meet its objective, based on previously set statistical considerations; and 2) The transfusion recipient septic transfusion reaction rate was close to double that of a parallel surveillance study on 5 day dated platelets conducted by the American Red Cross (yet with no statistically significant difference) raising safety concerns of the 7-day product. Subsequently seven day platelets were phased out over a 90 day period to minimize shortages and the blood banking system reverted back to storing platelets to 5 days.








Concurrent Studies


American Red Cross


The American Red Cross (ARC) supplies a large number of platelet products.ARC did not participate in the PASSPORT study, but used the BacT Alert device for screening platelet products stored for 5 days.†† The ARC uses a single aerobic bottle for sampling at 24 hours, does not re-culture outdated products and relies on following the septic transfusion rates from their platelets as a signal for product contamination rates.Before initiating the culture screening, the septic transfusion rate was 1/40,000.After initiating culturing of products, this rate decreased to 1/75,000. When diversion of the initial blood collection and doubling the culture volume from 4 to 8 mL were later instituted, the rate decreased to 1/140,000.


Septic reaction rate†††† before culture††††††††††††† after culture†††††††††††††††† after culture+ †††††††††††††††††††††††††††††††††††††††††††††††† ††††††††††††††††††††††††††††††††††† diversion + 8mL

††††††††††† (STR)††††††††††††† 1/40,000††††††††††††††††††††† 1/75,000††††††††††††††††††††† 1/140,000

††††††††††† N=†††††††††††††††††† ~500,000†††††††††††††††††††† 1,496,134††††††††††††††††††† 697,637


The ARC study also tracked the age of the platelets in relation to the septic transfusion rate.Day 4 platelets had a small increase compared to days 1-3 while day 5 had a substantial increase in the septic transfusion rate compared to earlier days.In addition, 3 mortalities were reported as associated with bacterially contaminated platelets and all occurred with day 5 platelets


Irish Blood Transfusion Service


In 2005 the Irish Blood Transfusion Service started culture testing of apheresis and pooled whole blood derived platelets with the BacT Alert device.Their protocol was to sample at 12 hours for apheresis and 36 hours for whole blood-derived platelets, inoculate 7.5 ml each into aerobic and anaerobic bottles.Platelets to be stored out to 7 days were re-sampled at day 4 with 7.5 ml samples inoculated into each an aerobic and anaerobic bottle and those that remained at outdate (day 7) were tested again.The confirmed positive results were as follows:


††††††††††††††††††††††† ††††††††††††††††††††††† Day 1 ††††††††††††††††††††††† Day 4 ††††††††††††††††††††††† ††††††††††† Day 7

Contamination rate ††† 1/3,300 (apheresis)††† 1/3,310 (all)†††† ††††††††††1/1,183 (all)

N=†††††† ††††††††††† 43,230†††††††††††††††††††††††† 3,310†††††††††††††††††††††††††† 8,282



Based on these results, the detection rate by the initial 24 hour culture was estimated to be 29% (False Negatives = 71%).





Lessons Learned from the Original PASSPORT Study


The sensitivity of the BacT Alert device in identifying contaminated products through a sample taken after the first 24 hours is poor (False Negatives approximately 74%).†† Slow-growing organisms do not proliferate to substantial levels until day 5 and possibly continue to proliferate over days 6 through 7.The PASSPORT study design did not track septic transfusion reactions by product shelf life day and thus it was not possible to determine risk based on storage age of the cells.Day 5 platelets were associated with a higher septic transfusion rate than day 3-4 platelets in the ARC study but this could not be correlated with the results of the PASSPORT study.The PASSPORT study did not have prospective stopping rules or a Data Safety Monitoring Board which would have made the decision to continue or halt the study much easier.




Platelet storage out to day 7 post collection remains desirable as a strategy to reduce platelet outdating and to provide greater flexibility for management of platelet donations and platelet inventories.For this reason there is a goal to reestablish conditions for use of 7 day platelets with added safeguards that would decrease the risk of septic transfusion reactions that is associated with longer storage.FDA seeks input from the Committee on conditions that would be adequate to address this risk.

Sponsor Proposal (Gambro/Fenwal/ AABB Task Force) on Bacterial Contamination


The sponsor has proposed to add new labeling to the platelet bags to warn the user of the results of the original PASSPORT study and to inform the user there are rapid bacterial detection tests that could be used to decrease the risk.Product labeling will specify use of standardized skin preparation and blood diversion methods and will direct users to sample the products at 24-36 hours, by inoculating 8 ml each into aerobic and anaerobic bottles.


The proposed revised PASSPORT study will consist of passive reporting of transfusion reactions. There will not be any cultures taken later in storage either prior to product issue or as surveillance cultures performed at outdate.The primary outcome of the study will be rate of sepsis which will also be tracked as a function of product age.There will be a Data Safety Monitoring Board to adjudicate and monitor the septic transfusion reaction rates, and predetermined stopping rules for the study will be defined.



FDA Proposal for Conditions of Use of 7 Day Platelets


FDA favors the addition of a second bacterial detection step during the course of platelet storage to reduce the risk of contaminated platelets on days 5 through 7.In brief, standardized collection methodology would be applied that includes a diversion pouch and standard skin prep. The products would be sampled on Day 1, by inoculation of 8 ml each into aerobic and anaerobic bottles.Products could be released after 24 hours (12 hours if there is a robust notification system for notifying transfusion centers with any positive results).Products remaining on the shelf on the morning of Day 5 would be re-sampled, again placing 8 ml each into aerobic and anaerobic bottles and could be released after 12 hours of culture if not positive.These products could be used for transfusion out to day 7.Outdated products would be sent back to the collection center and re-sampled placing 8 ml each into aerobic and anaerobic bottles to obtain the surveillance cultures.In addition, use of a rapid bacterial detection device would be required prior to release of products transfused at day 6 and 7. The septic transfusion rate will be tracked for each day of platelet storage including monitoring of the number of platelets of that storage age that are released. A Data Safety Monitoring Board will monitor the product contamination rates and septic reaction rates.There will be predetermined stopping rules for the study.


This approach would achieve several objectives:


a) Reduce the septic reaction rates associated with day 5 platelets (as reported in the ARC study) and reduce the risk of day 5 through 7 platelets which was the basis of ending the original PASSPORT study.


As demonstrated in the Irish study, testing at day 4 is not sensitive enough to identify contaminated products at day 7.Testing one day later is likely to increase this sensitivity.††


b) Validate performance of rapid bacterial detection devices.


Rapid bacterial detection devices have not been validated in a clinical setting and thus use of this device during the clinical trial would validate its clinical sensitivity and possibly justify its use as a stand-alone quality control or release test.



Questions to the Committee:


1.Does the Committee agree that to optimize culture sensitivity the initial sample taken at 24 hours of platelet storage should be tested with both an aerobic bottle and an anaerobic bottle using 8 ml volumes for each bottle?


2.Does the Committee concur with an industry proposal that the following collective measures are sufficient to allow use of 7 day platelets in the absence of a study to retest platelets for bacterial contamination on or after day 5 of storage?

  • Standardized antiseptic skin preparations
  • Sampling volume at 24-36 hours of 8 ml each into the aerobic and anaerobic bottle
  • Diverting the initial 30-50 ml of the initial collection away from the product
  • Labeling the bag to describe the risks of bacterial contamination identified in the original PASSPORT study
  • Reporting of septic transfusion reaction rate


If so, what would the committee consider as an acceptable septic transfusion rate for 7-day stored platelets?



3.If not,

††††††††††† a. Should an additional culture be performed on day 5? If so,

i) Would a single anaerobic culture on day 5 be sufficient?

ii) Should testing with a rapid bacterial test be additionally required on or after day 5?


b. Would rapid bacterial detection device on the day of issue on or after day 5 be sufficient, i.e. as a substitute for a day 5 culture?


c. Should the performance of the bacterial detection during the course of platelet storage (e.g. day 1 and day 5) be confirmed by a surveillance culture on outdated platelets (in addition to monitoring the rate of sepsis)?