November 2010
Feature Story

Anne Paxton

Eight years have passed since the CAP added an accreditation checklist item requiring that laboratories have a system in place to detect bacterial contamination of platelets; soon after, the American Association of Blood Banks adopted a bacterial detection requirement as well. It was a significant step forward in ensuring the safety of platelets.

But for whole-blood–derived platelets—estimated to be 10 percent to 15 percent of the transfused doses of platelets in the U.S.—there has always been a problem.

While apheresis platelets were easy to culture to meet the requirement, whole-blood–derived platelets, which are pooled from four to six donors, have been more difficult to culture. Surrogate methods of bacterial detection, such as pH and glucose tests, were just about the only game in town.

Now, however, the caliber of methods available for screening pooled whole-blood–derived platelets has improved enough that new standards are called for, and the use of surrogate methods could soon become a thing of the past.

Although many transfusion services may not yet have the news, beginning in June next year, it will no longer be acceptable in the CAP Laboratory Accreditation Program and in the BDP/BDP5 Surveys to use pH or glucose measurement to determine whether platelet units have been bacterially contaminated. An AABB standard that precludes surrogate testing methods will take effect Jan. 31, 2011. Apheresis platelets, which by some estimates now make up 82 percent to 90 percent of the market, are not affected by the new standard because they are usually cultured. It is only the potentially transfusible whole-blood–derived platelets that are affected.

This is the first major change in standards for detecting bacterial contamination in platelets since 2004, says James P. AuBuchon MD, advisor to the CAP Transfusion Medicine Resource Committee and president and CEO of the Puget Sound Blood Center, Seattle. “The College is changing in response to the recognition that what they call the surrogate methods for detecting bacteria, such as pH and glucose, and also microscopy, are really very poor, very insensitive means of performing that detection.”

“A lot of transfusion services don’t know this has been in the works,” Dr. AuBuchon says. “If they are paying attention to AABB, they would be getting an inkling of what is happening. We have put in the participant summary of the latest BDP proficiency test Survey for bacterial detection that a change in the CAP checklist is being considered. But there has been nothing formally out there yet, and we want to be able to give people a few months’ head start.”

The CAP Transfusion Medicine Resource Committee, which decided at its meeting last June to recommend to the CAP Checklist Committee an end to reliance on surrogate methods, has been discussing this move for several years, says committee member Mark Elliott Brecher, MD, chief medical officer of LabCorp, Burlington, NC. “What finally led us to put this into place is the fact that now, FDA-cleared systems are available to test whole-blood– derived platelet concentrates: the Verax Platelet PGD test and the Pall Corporation’s Acrodose PL System. Just a few years ago those were not available.” The FDA cleared the Acrodose PL System in October 2005; the PGD test was approved in November 2009.

The AABB made a similar decision and issued an interim standard in May 2010. “So since we’re basically on the same page, we thought it would make sense to try to sync up the two changes,” Dr. Brecher says.

The surrogate tests that have been used to detect bacteria in whole-blood–derived platelets fail on several scores, says Dr. AuBuchon, who is also the new president of the AABB. “One has to have somewhere in the range of 108 bacteria per mL, or thereabouts, to turn one of these tests positive. Furthermore, some bacteria will never turn the test positive. For example, the surrogate tests for pH or blood glucose are dependent on consumption of glucose by bacteria. Some bugs don’t consume glucose; they have other sources of energy, so the reagent will never detect them.”

Transfusion medicine specialists have known for some time that the surrogate tests are insensitive. With the surrogate methods, “you have to have a lot of bacteria in the bag—almost to the point where you can actually ‘see’ them—to detect any contamination,” says Zbigniew M. Szczepiorkowski, MD, PhD, a member of the Transfusion Medicine Resource Committee and director of the Transfusion Medicine Service of Dartmouth-Hitchcock Medical Center, Lebanon, NH.

There have been relatively few ways of dealing with that problem, Dr. AuBuchon says. While culturing is the primary means of detecting bacteria in apheresis platelets, and it’s the primary means by which blood centers document lack of bacteria, the culturing of individual whole-blood–derived units of platelets is difficult. “It’s very labor-intensive and consumes a lot of each unit.”

To help illustrate the importance of eliminating surrogate testing, the College has been running the bacterial infection BDP proficiency test for the past couple of years, “and in that Survey the participants who are culturing almost always get the right answer. But the sensitivity of the surrogate method is usually in the 50 percent range,” Dr. AuBuchon says. And that is despite the proficiency testing samples containing a large quantity of bacteria because the College wants labs to have a reasonable chance of getting the right answer.

For several years, the number of hospitals that have chosen to use only apheresis platelets has been increasing. At the University of North Carolina, where Dr. Brecher worked before LabCorp, as well as at Duke and Johns Hopkins universities, more than 99 percent of platelets are apheresis platelets.

At the same time, new safety measures, such as diverting the first 10 or 20 mLs of donations into a small pouch, have reaped benefits. The diversion pouch decreases the chance of contaminating the whole blood, and therefore the platelets, because the core of skin often taken by the needle during the blood draw, and its potential contaminants, may go into the pouch and not into the bag, Dr. Szczepiorkowski explains.

“If you look at apheresis platelets data from Johns Hopkins, for example, before culture was available,” Dr. Brecher says, “culture and diversion of the first couple mLs have resulted in transfusion reactions from bacterially contaminated platelets dropping 69.7 percent between 1987–1998 and the more recent period of 2004–2007. So we know that our interventions with apheresis platelets are effective, and we’re trying to make pooled whole-blood–derived concentrates similarly safe.”

Nationwide data show a trend toward greater use of apheresis platelets. The latest blood utilization data from the 2007 National Blood Collection and Utilization Survey Report indicate that 13,335,000 whole-blood platelet equivalents were collected in 2006, of which 10,939,000 were collected by apheresis and 2,396,000 were whole-blood–derived platelets (www.hhs.gov/ash/bloodsafety/2007nbcus_survey.pdf). That reflects a drop in collected whole-blood–derived platelets of 43 percent compared with 2004 and a 19.4 percent rise in collections of apheresis platelets. Transfusion data showed a smaller, but still significant, drop in whole-blood–derived platelet concentrates —down 15.7 percent in 2006 compared with 2004—while transfusion of apheresis platelets over the same period increased by nine percent.

Still, it may not be the case that whole-blood–derived platelets are being phased out. When Pall developed the Acrodose, the company was actually worried its market for whole-blood–derived platelets was going to disappear, Dr. AuBuchon points out. “That was because people were worried about bacteria, and there was no good way to detect bacteria in whole-blood–derived platelets.” Since then, however, “It does seem that the previously unrelenting charge toward apheresis platelet usage has stalled.” The most recent Department of Health and Human Services figures, reporting on 2006, show apheresis platelets made up about 82 percent of collections, and about 87 percent of transfusions.

The American Red Cross estimates that the risk of sepsis from platelet transfusion is currently about one in 175,000 bags. “That’s from both bacteria on the skin as well as in the patient’s blood,” Dr. Brecher says. “The data have always shown that when you culture platelets, two-thirds of the bacteria you culture tend to be gram-positive culture, such as you find on the skin. But the other side of the equation is which bacteria actually kill you, and it flips: It’s gram-negative organisms that we worry the most about.” Historically, in both the United States and Europe, these gram-negative organisms tend to account for two-thirds of the fatalities, he says.

Moreover, the organisms are much more likely to come from whole-blood–derived platelets, the American Red Cross also reports. Using comparable culture protocols, the organization found that bacterial contamination of whole-blood–derived platelets was 5.8 times the current rate for apheresis platelets (Benjamin RJ, et al. Transfusion. 2008;48:2348–2355).

Because of the culturing now available for whole-blood–derived plate­lets, that ratio is now something of a moving target, Dr. Brecher says. “As people have gone to better and better bacterial detection systems, we don’t know what the current difference is. There’s still definitely a difference, but it’s probably not as high.”

Data from a variety of sources, including from Johns Hopkins University and from Quebec and fatalities reported to the Food and Drug Administration, all suggest that interventions, at least with apheresis platelets, have lowered the sepsis and fatality rates probably by about 70 percent, Dr. Brecher estimates. “Some people would say that’s not good enough. But it’s a lot better than where we were a few years ago.”

Dr. Szczepiorkowski stresses that the upcoming change in the CAP accreditation requirement does not force a solution onto laboratories. “We did not say just one single solution; we are giving freedom to labs to decide what they want to do,” he points out. “But we want to be sure we don’t use an inadequate system to detect bacteria.”

Acrodose, manufactured by Pall and approved by the FDA in 2005, “allows for pooling of platelets early in storage, and then culturing the pool of platelets,” Dr. AuBuchon says. “This got around the FDA requirement to only have pooled platelets for four hours. It’s approved for five-day storage as pooled, and it allowed the application of more sensitive means of detection. It also got hospital transfusion services out of the difficulty of pooling platelets, something they didn’t like to do anyway.”

Acrodose was the only viable alternative until November 2009, when Verax Biomedical received FDA approval for its PGD test, which can be performed in roughly 30 minutes at a cost of about $25 to $30, about the same price that blood centers charge now for culturing. “This is a rapid immunoassay that is intended to be performed shortly before issue of whole-blood–derived platelets for transfusion,” Dr. AuBuchon says.

The options available to hospitals now using whole-blood–draw plate­lets, and wishing to continue using them, are essentially either to work with their blood centers and have the blood center implement the Acrodose system, or to continue to receive individual units of whole-blood– draw platelets and perform the Verax PGD test at the time the units are pulled for transfusion, Dr. AuBuchon says. Many large centralized transfusion services that use a lot of whole-blood–derived platelets, such as his own Puget Sound Blood Center, switched from surrogate testing methods to the PGD test several months ago.

Dartmouth-Hitchock uses the PGD system to supplement the cultures it performs. The PGD test resembles a pregnancy test, with two plastic windows displaying results for gram-positive bacteria and for gram-negative bacteria, Dr. Szczepiorkowski says. “The level of detection of the system is 104 for some organisms, and 105 for others, which is sufficient to protect patients from a major bacterial septic event,” he says.

The Centers for Medicare and Medicaid Services this year began considering tests for bacteria in platelets to be regulated analytes, which has added a new wrinkle to the bacterial contamination picture. “Generally, CMS does not regard a test used to qualify a blood component for transfusion as a regulated analyte,” Dr. AuBuchon points out. However, if a unit is found to be culture-positive, it’s possible that the donor of that unit might have to be notified of the finding before all the testing is completed, and the donor might have to be deferred. “Because of that potential deferral, even a temporary one, CMS said, ‘Well, that makes it a regulated analyte.’”

From the standpoint of CAP accreditation, there is also a complication of the switch to non-surrogate testing methods only: Proficiency testing is not yet available for the PGD system. “The reason is the PGD system works on detection of high level of colonies, so therefore you have to prepare the PT challenge in such a way you can actually mimic a real-life scenario,” Dr. Szczepiorkowski explains. “That’s different from regular bacteria contamination tests, which take a little volume of specimen, and the PT is in a small bottle that you put in an incubator.”

The PGD devices are more complicated, and that’s where the CAP Transfusion Medicine Resource Committee is working now to create PT, he says. It’s unlikely the proficiency test will be ready by Jan. 1, 2011, however, so each lab that uses the PGD system will have to come up with a twice-annual alternative assessment, possibly by exchanging specimens with another laboratory.

Platelets collected for transfusion might be contaminated with bacteria for two reasons, Dr. Szczepiorkowski notes. The primary reason is the area from which they are drawn was not properly cleaned, and a core of skin gets into the product and grows bacteria. The other reason is the person giving blood is bacteremic.

“That doesn’t happen too often,” he says, “but it’s why donor centers ask ‘Are you feeling well today?’” Other situations introduce bacteria into the bloodstream on a transient basis and mean that blood centers cannot completely eliminate the risk that a donor would be bacteremic. “Every time you go to the bathroom there are billions of bacteria in stool, and some of this gets into the bloodstream, but it’s quickly cleared from the blood in a question of minutes. Or someone may have an abscess somewhere and not realize it. Or, if you have cleaning of the teeth done, dental instruments can probe the gingiva and sometimes bacteria from the mouth can get into the bloodstream. Therefore, we don’t collect anyone who’s had dental work in the last 24 hours.”

Bacteria can enter the picture even if a donor looks perfectly healthy and the blood center does everything right, Dr. Szczepiorkowski says. “And I think that’s where the critical aspect is.” Usually the donor program will draw blood for a bacterial culture test 24 hours after collection. “That is good, but not perfect. It depends on how many bacterial organisms are in the bag. If they culture a couple of mLs, they may not draw the mLs that have the bacteria, so the test might be negative but the bacteria are still there.”

Dartmouth-Hitchcock’s donor center collects a bacterial culture after 48 hours, not 24. “After 48 hours the chances of detecting bacteria are higher because the bacteria grow and there will be more of them.” In certain situations the hospital has to use plate­lets before 48 hours, he adds, but the data show that the majority of severe reactions to platelets occur with platelets that are four to five days old. In such cases, “The bacteria has enough time to grow in enough concentration to cause problems.”

Confirming that a patient had a reaction from a platelet transfusion is not easy, but “whenever we have suspicion of a transfusion reaction, we do quite a thorough workup to be sure that, first of all, we find the problem, and second, we know what the bacteria was.” Usually if a patient develops sepsis and has chills and fever, “you culture the patient, and if you get the same bacteria as you get in the platelet,” Dr. Szczepiorkowski says, that is generally considered confirmatory. “If you have strep in the platelets and the patient grew E. coli, that’s obviously not the cause of the problem.” To prove it beyond a doubt, the laboratory can run a genotype of the platelets.

In the majority of cases, what the laboratory detects is skin contamination. “But there are some cases where the donor was, for example, a herpetologist or was playing with snakes and has some rare snake bacterium, or there are other cases of people with colon cancer, and some bacteria can actually enter their bloodstream. So it might be helpful in some circumstances to help the donor to identify the problem.”

At Dartmouth-Hitchcock, “We don’t test by culture until day two, and we also use the PGD system on day four, and that gives us some sense that even if the bacterial culture misses something, we still can detect it on day four.” In the seven years he’s been at Dartmouth-Hitchcock, “we haven’t had a transfusion-associated bacteremic septic event. Of course, people may say we lucked out, and that’s a possibility. But I think generally our number of platelets that are truly contaminated is about one per year. And in the vast majority of cases, we prevent those units from being transfused.”

There have been several false-positive PGDs at his center in the past three years, and he stresses that he is not suggesting people must do what his blood center does, a double system with a bacterial culture plus PGD. “However, if you have platelets that are six days old, which you officially can’t transfuse, if there is a huge shortage of platelets and a critical situation, I would at that point suggest using the PGD. I’d feel much more comfortable giving those platelets to people who require it in an institution where there’s a shortage. And shortages do happen in the U.S.”

Improvement in phlebotomy practices may be having an effect. “People have switched to iodine and other solutions, which actually makes their disinfecting better,” Dr. Szczepiorkowski says. “And we’ve recognized that skin is a very complex tissue with a lot of little crevices and holes where bacteria can stay. So the diversion pouch is very important, and appropriate timing and use of appropriate systems to disinfect the skin are critical. We have to have a very safe system of collection—then detection is just the next step.”

Most laboratories have had some warning of the upcoming ban on surrogate methods for detecting bacteria in platelets, or at least shouldn’t be surprised. “If the lab is AABB-accredited, of course, they have known about this before. And the notes in previous CAP checklists have quite clearly stated glucose and pH are really not good methods,” Dr. Szczepiorkowski says. “I think there’s awareness now, and the last several years of PT have also shown those methods just don’t work.”

“The Transfusion Medicine Resource Committee did want a concerted effort to make sure CAP members are aware of the change, although if they buy products from the outside, and they’re apheresis platelets, it really doesn’t affect them at all. They just need to ensure that their platelets provider uses appropriate systems, to know what the systems are, and to make sure that’s part of their contract. The vast majority have that contract with their blood supplier.”

If there is a septic bacterial event related to platelets at a hospital, the laboratory is in a sense responsible, he says. “So it’s very important for them to understand what’s being done and not have to stand up one day and say ‘I didn’t know that.’”

Managing the transition to the new requirements will ensure the success of essential improvements in the safety of platelets. “There’s no question whole-blood–derived platelets are necessary,” Dr. Szczepiorkowski says. “We need them. But there has to be negotiation between the transfusion centers and donor centers to be sure the platelets received from the blood centers are safe.”