October 2002
Cover Story
Anne Paxton
In the late ’90s, it was called "exquisitely sensitive" and hailed as a silver bullet in the nation’s long campaign to end the risk of contracting HIV or HCV through blood transfusion. Nucleic acid amplification testing looked as close as we could get to a test that would close the window of viral transmission from donated blood—or at least drastically decrease the risk of transmission.
In fact, that turns out to be the trouble.
Rapidly adopted in Europe for plasma derivatives, NAT took root in the United States surprisingly quickly for all donations destined to become transfusible components, despite its cost and immense logistical difficulties. It became a routine part of blood bank testing in 1999, even though it was still in clinical trials and licensing applications for the assays were awaiting Food and Drug Administration approval.
Today, NAT testing for HIV and HCV is performed on nearly 100 percent of U.S. blood donations, and clinical trials have borne out the blood services industry’s hopes for the technology. Over 24 million samples had been NAT-tested for HIV and HCV as of November 2001. NAT caught 88 specimens that were positive for HCV RNA, and seven that tested positive for HIV-1 RNA.
"Clearly, no one in their right mind would be transfusing blood that hasn’t been NAT-tested," says Ronald Sacher, MD, director of the Hoxworth Blood Center at the University of Cincinnati and a member of CAP’s Transfusion Medicine Resource Committee. "It’s the best test we have, and there are no other tests that could in fact be as definitive in terms of shortening the window period of infectivity."
Chiron Corp., which distributes the Procleix System developed by Gen-Probe Inc., already has 75 percent of the NAT-testing market for whole blood. So when the Procleix System became the first to win approval from the FDA last February for testing of whole blood donations, it may have seemed anticlimactic. But licensing marks a milestone for NAT testing. It means blood centers are effectively required to perform NAT testing of blood donations (using either Chiron’s test or that of Roche Diagnostics, which is awaiting FDA approval).
Licensing also means blood centers can now label their products NAT-tested, manufacturers can make a profit on their NAT tests, and blood centers will pay a lot more for the assays. "Frankly, the primary impact of licensing is cost," says Jim MacPherson, chief executive officer of America’s Blood Centers. "When it is a research project, the FDA requires that the manufacturers can only recover their costs. They can’t make a profit or recoup their investment; they can only charge exactly what the test costs. Now they can charge a profit."
Furthermore, NAT testing will become more entrenched in the blood industry. "For the foreseeable future, we’re going to be using NAT for any direct virus test we use," MacPherson says.
NAT testing is tailor-made for easing public anxiety over the blood supply. "With NAT testing, we’ve been able to say to patients that, due to advances, the risk of HIV when receiving banked blood is almost one in 2 million," says Katharine Downes, MD, associate director of blood banking and transfusion medicine at University Hospitals of Cleveland and a member of the College’s Transfusion Medicine Resource Committee. "We’ve been able to allay a lot of fears. But the public’s expectation for a zero-risk blood supply persists."
James AuBuchon, MD, vice-chair of the CAP committee and professor and chairman of pathology at Dartmouth-Hitchcock Medical Center, Lebanon, NH, specializes in decision analyses and cost-effectiveness issues. He summarizes the dilemma posed by NAT: "As the blood supply gets safer, there are fewer cases of transmission remaining to be prevented and therefore the cost-effectiveness of continued attempts to improve safety will be very poor."
"It’s certainly been true in the case of NAT," Dr. AuBuchon continues, "and we will face the issue again when we consider moving from minipool NAT to single-donation NAT. These steps will appear very desirable to many. However, the morbidity and mortality caused by the residual number of transmissions either before or after minipool NAT are orders of magnitude smaller than other risks that patients face in transfusion."
Blood centers have been screening for HIV since 1985 and for HCV since 1990. But pressure to find more precise probes has never let up.
Blood centers actually used NAT tests under extended investigational new drug exemptions with the FDA, says Neil Gunn, vice president for marketing with Chiron Blood Testing. "The FDA considers it an extremely important technology and it allowed extended INDs to be carried out under a cost-recovery program."
Shrinking the window
NAT’s advantage is that it detects viral genes rather than antibodies or antigens (proteins for the virus). As a result, it can find viruses earlier. "Surface antigen takes a period of time to develop. The beauty of NAT is it doesn’t wait. It actually searches directly for the viral RNA or DNA, so you can detect its presence much earlier," Gunn says.
While Chiron uses a process called transcription-mediated amplification for its NAT testing, and Roche uses polymerase chain reaction, "at the end of the day the two tests are identical," MacPherson says
In fact, NAT testing has slashed the window between viral infection and the ability of an assay to detect the virus. The window for HCV fell from 82 to 25 days, and the average HIV-1 window fell to 12 (compared with a 16-day window with antigen testing).
"The biggest advantage of this test has been with hepatitis C, as one would predict because of its longer window period," says Laurence Sherman, MD, JD, co-chair of the Transfusion Medicine ResourceCommittee and professor emeritus of pathology at Northwestern Medical School in Chicago. "The impact was less for HIV, because with the most recent tests that had gone into use for HIV, one was already looking directly for the virus rather than just the antibody."
But those tests will soon recede into the background. "The approval of the Chiron/Gen-Probe test allows the removal of the requirement for p24 testing for HIV," Gunn says, noting that it is one of the unique features of the FDA’s licensing of Chiron’s NAT test. "Under our package insert, blood centers could actually eliminate p24 testing for HIV. The major rationale, the FDA said, is that the NAT assays should be able to identify all known HIV subtypes."
There is pressure from the American Red Cross and America’s Blood Centers to move forward on eliminating p24, Dr. Sherman says: "There’s been a lot of interest in doing that, particularly since the cost of blood products has risen much more rapidly than other things in the health care arena—let alone the CPI. From a laboratory standpoint, the increase in the cost of blood products is much greater than anything else."
The p24 antigen phase-out could at least trim possible enormous costs for blood centers as NAT testing widens its scope. "Our hospitals don’t like it, of course," Buff Mair, MD, medical director of Florida Blood Services, St. Petersburg, and a member of the CAP transfusion medicine committee, says. "We try to absorb the extra costs as much as possible, but ultimately as the test gets more sophisticated and more expensive, it gets passed on to the consumer—which is the hospital."
Most blood centers expect that Roche Diagnostics’ NAT tests for HIV and HCV will also get FDA approval next year. "We are on a fast track," says Patrick Aquilino, Roche’s marketing manager for blood screening. The Roche assays are used to screen about 25 percent of the U.S. blood supply. "We are making every effort to facilitate approval of our applications with the FDA," Aquilino says.
Ironically, though only Chiron will be able to make a profit for the time being, blood centers won’t be able to use the Roche test to save money. "Roche has been charging a royalty ever since it settled its patent dispute with Chiron, and the royalty went up significantly after the Chiron test was licensed," MacPherson explains. "And although that is a profit to Chiron, it is a cost to Roche."
Licensing will also increase the impact of NAT testing on blood safety. Although U.S. blood centers almost universally performed NAT testing while Chiron’s license was pending, they did so under "Phase 1" conditions of the investigational new drug exemption—which meant that blood components were sometimes not quarantined. They could be issued, and potentially transfused, before the NAT results became available.
Michael P. Busch, MD, PhD, professor of laboratory medicine and vice president of research and scientific affairs for Blood Systems, the second largest blood collection organization in the U.S., reported in 2000 that each positive donation could well have infected two to three recipients, but, as it turned out, only one NAT-positive component was transfused.
Dr. Sherman notes that in addition to the lack of quarantine for new blood components, often frozen products such as plasma and cryoprecipitate that were drawn before blood centers or hospitals started NAT testing, and that were still on their shelves, weren’t exchanged for test-negative products. "In late 1999, at an FDA workshop the majority of institutions had not exchanged frozen products," he says.
The FDA had set a target date of mid-2000 for everyone to treat NAT tests the same as licensed tests, even though they couldn’t label it, Dr. Sherman adds. He believes very few people have been transfused with these products, and that in fact the pre-existing systems and tests were quite good. "The issue really was if you’re doing it, and charging for it—and virtually everybody is—then the people you’re charging ought to be getting a test-negative product," Dr. Sherman says.
The CAP Transfusion Medicine Resource Committee only recently completed plans for a proficiency test for NAT, which is scheduled for introduction early in 2003. "This is not an easy test," Dr. Sherman says. "CAP’s microbiology committee has been doing a proficiency Survey for viral load NAT testing—measuring how many HIV virions are in a person who’s undergoing treatment—for some time. And it’s been apparent from the proficiency tests there that the tests aren’t as precise as one would like. We spent a fair amount of time developing a Survey for blood donor centers proficiency testing for NAT as well. It really took longer than we expected, but we finally have it and it’s in the catalog."
One practical problem of developing a proficiency test, says Dr. Sherman, was that "CAP to the greatest extent possible does not ship infectious material, unless we’re sending something to see if you can grow a bacteria."
Policy and plasma
NAT testing’s widespread use in whole blood has been closely tied to its use in the plasma industry—even though different methods and technologies are used to perform the assays. Laboratory Corporation of America Holdings was first to be FDA licensed for NAT testing. LabCorp’s test is a screening test for plasma, which differs from whole blood in several major ways. Plasma has a number of different uses, including its use as a raw material for the manufacture of clotting factors for hemophiliacs and immune globulin, whereas whole blood is used primarily for surgeries and emergencies in which large quantities of blood cells are lost. The key difference is perishability.
Andrew Conrad, PhD, chief scientific officer of LabCorp’s National Genetics Institute, notes that "plasma is very, very stable, but whole blood isn’t." Whole blood is good for a few weeks and then discarded, whereas plasma can remain frozen for use for years. The whole blood tests are done quickly on individual samples, or in most cases small pools, to make the whole blood available as soon as possible. "Plasma screening tests are done on larger pools," Dr. Conrad says, because plasma collections can be held in storage, allowing for more complex resolution algorithms that take more time and allow for larger pools.
"A small blood bank could wait weeks to get pools of more than 500 samples," Dr. Conrad says, making the whole blood essentially unusable by the time it gets tested. Plasma, however, can be frozen and held as long as needed to accumulate the desired number of samples for testing. The basic ideas for testing whole blood and plasma are similar, but, Dr. Conrad notes, "plasma has a bit of a luxury in that the time pressures aren’t the same." Economies of scale are another advantage deriving from plasma’s lower perishability. Whole blood tests need a system that can be put out in many locations, whereas a central plasma testing laboratory can get huge shipments in to test.
NAT tests for plasma, in fact, are sometimes used as confirmatory tests when NAT tests of whole blood are positive. For all whole blood that uses NAT tests, a positive result is confirmed with an independent, different test.
How can the plasma test be used for whole blood? Once the test result on whole blood is positive, the blood cannot be used—so time is no longer of the essence.
In addition, although the blood cells in whole blood are perishable, the whole blood can be spun and the plasma recovered. "Here is where the two worlds meet," Dr. Conrad says.
Steven H. Kleinman, MD, a pathologist with the University of British Columbia in Vancouver, noted at a recent FDA Blood Products Advisory Committee meeting that the source plasma industry sets a standard and influences the whole blood industry, in a way that has trumped FDA regulation.
Indeed, Dr. Sherman points out, the European Union’s requirement that all plasma shipped there be NAT-tested jump-started the near-universal adoption of the technology in the United States. "It was part of the reason for getting it up and running. A lot of blood centers here ship plasma to Europe, and the EU was not accepting any plasma that wasn’t NAT-tested after July 1, 1999," Dr. Sherman says.
"The European Union was quicker to jump on this requirement," Dr. Sacher agrees, although he downplays the EU’s requirement as an incentive for U.S. blood centers to adopt NAT testing. He notes that part of the EU’s alacrity stemmed from the inactivity of many European blood supply officials in the early part of the AIDS epidemic; in some countries directors of testing were criminally indicted.
However, Europe has not wholeheartedly embraced NAT for whole blood donations. While Germany and France require it, the United Kingdom currently only tests for HCV with NAT. "It does not conduct HIV NAT testing because it believes that the cases it would find are relatively low in number," Chiron’s Gunn says.
The UK is not alone in its concern about the price of NAT testing. There has been a huge uproar over pricing in Europe, and it has spawned major litigation, Dr. Busch notes: "The Europeans developed these assays themselves and brought them up before we did; then Roche and Chiron are saying in addition to $1 or $2 per donation, give us $5."
Gunn suggests that NAT testing is worth the extra cost, in part because it can improve both the safety and the availability of blood products. "All too frequently there are blood shortages, and we want to work with the blood industry to help reduce the impact of these," Gunn says.
Competing patents
A web of patent rights has complicated the availability and development of NAT testing. Roche Diagnostics has a patent on the PCR technology used for nucleic acid testing. "That’s why Gen-Probe/ Chiron had to come up with something different and it developed the TMA approach," Dr. AuBuchon says.
However, Chiron, which cloned and first identified HCV as the cause of transfusion-related non-A, non-B hepatitis, as it was then called, has granted nonexclusive licenses to its HCV technology to many of the large diagnostics companies. It recently settled a patent dispute with Roche over this technology with an agreement that Roche would pay tens of millions of dollars to Chiron and pass through future royalties on the HCV test to Roche’s customers.
Essentially, this means that Chiron has a monopoly on NAT testing for HCV, Dr. Busch explains, and it licenses the technology to other companies such as Roche. Chiron charges a royalty per donation even though the samples are pooled for testing—which Dr. Busch considers a pricing structure that artificially "skews the potential decision to move the nation from pooled to individual-donation testing."
"They are charging a $5 royalty no matter whether the test samples are in pools of 100s or singles," Dr. Busch says. "As a result, moving from minipool to single-unit testing doesn’t drive prices up dramatically. It may only increase them marginally." He predicts that costs will remain high because of such patent issues. "To my mind, we’ve seen prices go up progressively more than anticipated, so that’s the end of discussion of the so-called economy concept that with multiplexed assays you’d save money."
Despite these factors, competition in NAT testing survives. "We worked very, very hard to make sure both companies stayed in business," says MacPherson. "We did not want to see Roche forced out because of the patent dispute, so we encouraged both companies to settle, and they did." Competition between companies is almost always good, he believes, because it stimulates improvement. In this case, "Chironthrough its partner Gen-Probe brings innovation, but we also know that Roche has the world market in NAT testing, and we wanted both," MacPherson says.
Drs. Busch and AuBuchon stress that to some extent the attention focused on NAT could profitably shift to other, more cost-effective measures. Dr. AuBuchon says, "The two major risks that transfusion recipients face today are incompatibility, which in red cells causes acute hemolytic transfusion reaction, and, for platelet recipients, bacterial contamination. Each of these appears to cause between one and two dozen deaths in the U.S. annually—far outpacing the frequency of HIV [one per 2 million]."
Individual-donation vs. pooled testing
The most contentious debate is now over individual-donation testing, which is being seriously considered because, despite the sensitivity of NAT, a few infections still escape detection when samples are pooled.
"The issue is really one of dilution," Dr. AuBuchon says. "In a minipool as currently set up in the U.S., between 16 and 24 units are tested simultaneously. Statistically, only one is going to be an infected unit. Therefore HIV in that single hot sample is going to be diluted by the negative members of the pool, and if the concentration of the genome is near the detection level, it may be diluted below the detection level."
That possibility is more than theoretical. At a Sept. 12 meeting of the Blood Products Advisory Committee, the FDA announced that three cases of HIV transmission during the window period have occurred since the introduction of NAT. Indira Hewlett, PhD, director of emerging and transfusion-related diseases at the FDA, said at the meeting, "It is FDA’s current intention that when feasible and when the appropriate platforms become available, ID [individual donation]-NAT would eventually be implemented for mass screening of donors."
"Things are still very dynamic," says Dr. Busch. "Unfortunately there have been a couple of breakthrough transmissions where the minipool actually didn’t detect an infectious donor." As a result, some blood centers are going to ID testing—a trend driven mostly by the military, which tests only 40 to 50 samples a day in its laboratories. The Oklahoma Blood Center, a frequent early adopter of new screening technology, has also announced it will do individual-donation NAT testing.
"Our testing site is telling us if an individual center said, ’Look, we want individual NAT on our donations,’ they could actually do that, and be charged accordingly," says Francis R. Rodwig Jr., MD, chairman of pathology and laboratory medicine at the Ochsner Clinic Foundation in New Orleans and member of the CAP transfusion committee. "They’re still working the price out, but it would be a lot more expensive, and it’s a big concern because we know there’s a lag between increased charges and increased reimbursement."
Dr. Busch is similarly skeptical because of the extremely high cost of ID testing compared with the gains. "The relative benefit of ID-NAT when compared with minipool NAT is very small because the ramp-up phase of viremia is logarithmic, with a double time of 21.5 hours," he says.
Urgency of automation
While the American Red Cross has said it is considering an intermediate reduction of pool size, which would partially address the dilution problem, most blood center officials as well as the FDA consider ID unfeasible without automation—and the industry is only halfway there.
Manufacturers have improved on the traditional PCR testing laboratory, Dr. Rodwig says: "Traditional PCR was a person sitting in a laboratory with a pipette, doing hours and hours of delicate testing. Now the samples are collected and pooled, they go through an analyzer, and the results are spit out at the end."
But fears of cross-contamination, possibly causing the wrong DNA to be amplified, make any process of PCR difficult to automate. Referring to the extraction steps in the Roche test and the extensive manual pipetting required for the amplification phase of the Chiron test, Celso Bianco, MD, executive vice president of America’s Blood Centers, urged Gen-Probe, Chiron, and Roche to hasten their development of automated instrumentation.
Chiron’s test is a semiautomated system, Gunn says. "A large amount of it is automated but some of the individual steps are quite manual, and we are looking to automate those as we go forward over the next couple of years, with a view toward full automation in the mid-term."
Chiron has been promoting single-unit testing to small labs, Dr. Busch says, because it can give them a competitive advantage over large labs: "Thelarger megalabs like ours or the Red Cross clearly can’t go to single-unit until there’s appropriate automation."
"It’s not so much a cost issue as having to increase staff three- or fourfold, when most can’t even hire enough staff to keep afloat now," says Dr. Busch. "The companies themselves would be the first to admit that anyone that tests over 100,000 donations a year would be pushed to the edge if they tried to turn around test results in a timely fashion."
Dr. Mair agrees. "The logistics of single-donation testing are complicated because NAT testing is not automated." At Florida Blood Services, which performs 825,000 NAT tests per year, one technologist can process 2,816 samples with eight to 10 hours turnaround time, if they are negative; positive tests require additional testing. "It does require hands-on tech time, and there’s a nationwide shortage of techs," she says.
Clinical advantages weighed
Hoxworth Blood Center’s Dr. Sacher says three cases have been reported in the past yearin which the NAT test was negative but HIV was transmitted. "So even though the window period for HIV is probably now about six to seven days, there’s still the potential of a donor who has the virus in their system not being detected by NAT," he says.
Eventually, MacPherson believes, blood centers will go to single-donor NAT. "There’s a pretty small clinical advantage to single-donation testing for HCV. You’ll only close the window by a couple of days because the virus ramp-up is so fast," he says. "With HIV you can close the window maybe another seven to 10 days by going to individual testing. The controversy is whether it’s worth it to multiply the workload, let alone the cost, by 16 times to pick up one case of HIV per year."
Dr. Busch is skeptical the tradeoff is warranted: "Our analyses would suggest we’d probably pick up three additional HIV- and three additional HCV-infected donors per year in this country, out of 13 million donors, that are currently missed by minipool and could be detected by single-unit. The problem is it will cost us $100 million to pick thoseup."
The FDA said in September that the currently licensed platform "requires upgrades including automated sample preparation, reagent preparation, etc., to maximize its efficiency for high volume use for ID-NAT. These upgrades and the necessary regulatory submissions for their approval will require time. Therefore, although ID-NAT may be technically feasible, it may not be operationally so at the present time."
The question of hepatitis B NAT testing has generated similar debate. There is no HBV NAT test yet in the United States, but this past summer Roche began clinical trials of its Cobas Ampliscreen HBV test. The Japanese Red Cross is already performing HBV NAT testing to screen blood donations as part of Roche’s triplex AmpliNAT assay. Chiron has also developed a triplex test, set to begin clinical trials in Europe in 2003, that looks for HBV as well as HCV and HIV.
The market is unquestionably there. "Around the world there are probably 300 million people with hepatitis B," Gunn points out. "The demand for an HBV NAT test will come from its prevalence within the blood donor population, and we’ll get a much better estimation of that in 2003." Antigen tests already in place are less sensitive than NAT would be, he adds, and preliminary data suggest the HBV window can be reduced by 20 to 30 days over current assays.
However, Dr. AuBuchon says, "Hepatitis B causes serious consequences or long-term chronic complications much less frequently than HCV. So the advantage of adding the HBV primers to NAT will be relatively limited." In addition, the low level of viremia associated with HBV at the beginning of infection is a related drawback of NAT testing for the virus.
Says Dr. Downes: "There is also a very slow doubling rate of HBV, about every three days, which makes the amount of HBV DNA present in the pooled sample much lower than in HIV and HCV in the early part of infection. So performing NAT on HBV pooled samples appears to be less useful."
Other NAT tests likely to be adopted are for hepatitis A and parvovirus B19, although there is little consensus that they are urgently needed for blood safety. "The reality is people who collect whole blood are having to do things that don’t make sense in the whole blood sector, and do make sense in the source plasma sector, and it may be because they want to be competitive in the market for their fractionated products," Dr. Kleinman says. "I think everybody should recognize that because of safety concerns in the source plasma industry, the whole blood industry has implemented a series of tests, some of which would have been implemented anyway, like HCV and HIV NAT perhaps, but others of which, like parvovirus B19 and HAV NAT, probably have no value for blood donors and probably no value for the final products."
Stemming a new threat
West Nile virus is the latest potential candidate for detection by NAT assay. In September, the Centers for Disease Control and Prevention confirmed that four recipients of organ transplants from a single donor had contracted West Nile. The organ donor, who also turned out to be positive, had been transfused with blood from more than 60 blood donors.
West Nile has been a particular concern in Louisiana, which has had a large number of cases and deaths. Ochsner Clinic Foundation’s Dr. Rodwig says there has been a lot of uncertainty about testing because there has never been a case definitively associated with blood transfusion. "It would be impossible to quickly develop a NAT test," he says. "Of course you could do an antibody test—but we don’t know if that would just unnecessarily screen out donors."
The transfusion link is still theoretical, but the CDC and FDA said Sept. 19 that West Nile virus can survive in some blood products and can "probably" be passed through blood transfusions.
As a result, federal officials are increasingly convinced that they need to quickly develop a screening test to protect supplies in blood banks. NAT testing, in fact, will be well suited for West Nile if it proves cost-effective. Since peak viremia occurs one or two days before the onset of illness, antibody tests would not be useful as blood donor screens.
Public alarm over the polio-like effects of West Nile all but ensures that NAT testing for the virus will eventually be required. And many blood experts like Dr. AuBuchon caution that perspective, not alarm, is needed to keep blood policy cost-effective.
"NAT testing is a versatile technology that will allow us to expand our oversight of the blood supply against new threats," Dr. AuBuchon says. "But I would not want us to forget that there are other, much larger problems than viral transmission."
Anne Paxton is a writer in Seattle