HIV, HCV windows nearly shut with minipool NAT
Fast, functional, foolproof. The quest for all three traits
in a screening test for HIV-1 and HCV continues. But a new study of three years’
worth of U.S. blood donations gives the first proof that minipool nucleic acid-amplification
testing, or NAT, has brought blood centers closer to their goal.
As reported in the Aug. 19 issue of the New England Journal of Medicine
(Stramer SL, et al. 2004;351:760-768), the study found that the use of mini
pool NAT prevented about five HIV-1 infections and 56 HCV infections per year.
After all of the nation’s major blood centers adopted NAT in 1999—accounting for more than 98 percent of blood donations—the residual risk of a transfusion transmission of one of the viruses dropped to one in 2 million blood units, concluded the study, which was funded by the National Heart, Lung, and Blood Institute.
The study group used NAT to analyze all antibody-nonreactive blood donations between March 1999 and April 2002. Twelve of 37.2 million seronegative donations were found to be positive for HIV-1 RNA, while 170 of 39.7 million seronegative donations were found to be positive for HCV RNA.
Two more years of data from the American Red Cross allowed the study group to evaluate also the dynamics of HIV-1 and HCV seroconversion.
Blood centers knew that NAT could shrink the window between infection and detection, says lead author Susan Stramer, PhD, the American Red Cross liaison to the CAP’s Transfusion Medicine Resource Committee.
But there were two problems in 1998: NAT was not yet approved by the Food and Drug Administration, and it was impractical to perform NAT on individual donations.
"The American Red Cross, like many other blood collectors and providers, became interested in NAT as a way to close the infectious window period that remains for several agents for which we do testing, and for which there are concerns about residual rates of transmission," says Dr. Stramer, executive science officer at the ARC’s National Testing and Reference Laboratories, Gaithersburg, Md.
"The technology was out there—PCR and transcription-mediated amplification—and the FDA was very supportive of advancing these technologies in the blood screening environment."
In fact, former FDA commissioner David Kessler, MD, in a 1994 workshop, had challenged the laboratory testing industry to come up with creative ways to close the HIV window. Manufacturers like Gen-Probe were working on basic NAT research under grants from the NHLBI.
Starting in 1998, "We had several meetings with the FDA, the manufacturers, and a task force from the American Association of Blood Banks to work out a testing algorithm. A lot of hurdles were overcome so that we could implement testing initially at the pre-approval IND [investigational new drug] stage."
There was no precedent for starting to implement the test before FDA approval, she says.
"In a conventional clinical trial method, you test 10,000 donations, then stop, the FDA reviews the data, and so on. But here that would not be ethically sound, because we had the knowledge that we were interdicting infectious units. Not only the recipients but also the donors have to be told of the results so they don’t transmit an infection."
But the major logistical barrier was not lack of an approved test, or even cost. It was NAT’s turnaround time.
"We don’t have a system in place to test each individual donation in a reasonable time period. NAT is very labor-intensive," Dr. Stramer notes.
At the time, the plasma industry was already beginning to require NAT. "But the pressures in whole blood are very different," she says, "since testing has to be done quickly enough so time-sensitive cellular products can be released." Lacking FDA approval, NAT could not yet replace conventional antibody and antigen screening, so it had to be done as a supplement to other routine testing.
"Using minipools was the only way we could implement testing in the time we did," Dr. Stramer says.
With plasma, there will typically be 512 donations in a pool, she explains. "For whole blood, we have smaller pools, ranging in size from 16 to 24." The two nucleic acid-amplification tests employed during the clinical trials, and approved by the FDA in 2002, use minipools: the Gen-Probe Transcription-Mediated Amplification system with a minipool of 16 donor samples, and the Roche Molecular Systems Cobas AmpliScreen HIV-1 and HCV tests with mini pools of 24 samples.
"Minipool testing is very effective for HIV-1 and HCV because the virus in those agents replicates very quickly once you’re infected. The time difference between infection with the virus and when it’s detectable in pools is very, very short," Dr. Stramer says.
In addition to demonstrating NAT’s overall value, the study
showed that NAT testing is better than HIV-1 p24 antigen testing at detecting
infection in the window period.
In fact, NAT is anywhere from five to 10 times as effective as p24 antigen, Dr. Stramer says. "It’s hard to believe that a test done on a pool basis is still more sensitive than a test for p24 antigen. But during the clinical trial, we collected all this data, and additional studies done by the manufacturers also demonstrated that p24 was actually redundant, and considerably less sensitive even when doing these small minipools."
A very small number of units that were interdicted during the study period had other marker activity that would have prevented their being released for transfusion, she says. For example, for HIV-1, two of 12 RNA-positive units had p24 antigen activity. About one-third of the HCV RNA-positive donations would also have been deemed nontransfusable because of an elevated alanine aminotransferase level or reactivity to other routine screening tests.
Another study finding was that the respective rates of positive HCV and HIV-1 RNA tests were 3.3 and 4.1 times higher among first-time donors as among those who donated repeatedly. This was not surprising: "First-time donors are more generally representative of the U.S. population than repeat donors," Dr. Stramer notes.
With the ARC followup data, the study group demonstrated or corroborated previous literature that showed that an HCV-infected individual has a number of different outcomes, she says.
"We know most people with HCV will become chronic carriers. Only about 20 percent of individuals infected will clear the virus, and we basically confirmed the findings in other populations. It’s important to understand the differences between donors who clear and those who don’t."
Interestingly, the study found a small number of donors—three—who were "immunologically silent"; they were infected but for some reason never produced an antibody response. "Even though they had high-titer viremia, their own host defenses never recognized the virus as foreign. They were perfectly healthy. They just coexisted with a high-level viremia."
The mechanism isn’t understood, she says, and how long it would persist is unknown. "There’s no defect in antibody production, so perhaps it could be cellular immunity. But it’s reassuring that the number we find is very few."
Despite the proof of its worth, there is no evading the fact that NAT is phenomenally expensive, Dr. Stramer says.
Two independent groups determined that using minipool NAT for HIV-1 and HCV testing, even with the elimination of p24 antigen screening, cost $1.5 million to $4.3 million per quality-adjusted year of life—a price tag that dwarfs that of many other public health measures.
"NAT is more cost-effective for HIV because HIV causes incurable disease," she says. But overall, NAT is very cost-ineffective—mainly because the risk of infection from blood transfusion is far outpaced by other risks of being treated by the health care system. "The price is high, and our yield is low because for the most part blood donors are very safe."
Using minipools doesn’t bring sizable financial savings. "As it turns out, manufacturers are charging us a per-donation price. From their standpoint, they’re developing a very expensive technology and they want to recover all their R & D and production costs," Dr. Stramer says.
"If they allowed us to do pool testing on a per-test basis, they would be making 16 to 24 times less profit, so they’re charging us per donation, and there will be an incremental charge if we move to single-unit testing."
It is largely the same manufacturers that offer NAT testing worldwide. "In Asia, Europe, and North America, we’re all doing NAT the same way. Most large blood centers are using minipool testing," she says.
"Some smaller centers in countries with very decentralized testing, like Italy, all do single-unit testing. For them, it’s not efficient to do pooling, because with limited donations in any given day it’s simpler to do the test as the donation comes in."But the technology and effectiveness of testing are no different in the U.S. than in Australia, Japan, or Europe, she says.
Minipool NAT is less appropriate for hepatitis B screening.
"With HBV, the NAT testing questions are a little different. The yield is very comparable to HCV in clinical trials, but there’s a lot of debate on whether HBV testing in minipools is warranted relative to waiting for single-unit testing," Dr. Stramer says.
Early infection with HBV is much more prolonged and the virus doesn’t replicate as quickly, so the yield is far higher with single-unit relative to minipool testing, she says.
"We also have another good test, which is HBV surface antigen, and as we get tests with improved sensitivity for the antigen, it diminishes the yield of minipool NAT. In the case of HIV, antigen testing wasn’t very good, but it is good in HBV, and it’s probably going to be better as new tests are introduced."
In August, the FDA’s Advisory Committee on Blood Safety and Availability recommended against requiring blood centers to use a new HBV nucleic acid-amplification test developed by Roche.
The most urgent research need in NAT, Dr. Stramer suggests,
is probably automation of tests. Roche and Gen-Probe are aggressively working
on automation, though it won’t come without an extremely high cost.
A walkaway technology with tight process control and limited user intervention will be needed to facilitate a shift to individual testing, she says. "As automation progresses, we will continue to evaluate and monitor the situation and determine when to transition to a smaller pool size or eventual single-unit testing."
Any automation will need to be sufficiently flexible to accommodate screening for emerging path o gens. As it happens, NAT’s cost-effective ness is improving because of the new pathogens it can screen for, such as West Nile virus. In 2003, after NAT was implemented to screen for West Nile, nearly 1,000 blood donations were identified and discarded before transfusion.
"The knowledge we had about the process—and the fact that we had already implemented such a pooled test for HIV-1/HCV—allowed us to implement another agent using the same platform and same processes rather quickly."
"We’ve done all the legwork up front," Dr. Stramer says. "We knew how to create pools, verify sample addition, test pools, how to manage reactive pools, resolution testing, donor management-all the things involved were already worked out."
"So when West Nile was identified as a transmissible agent in September 2002," she says, "for us then to implement testing nine months later was only possible because we had the infrastructure in place, the test kit manufacturers had reagents available, and the FDA was supportive of the national IND testing model."
True to prediction, the study showed that NAT is effective in improving detection of HIV-1 and HCV infections. "This project also demonstrated," she adds, "that with cooperation between blood pro vid ers, manufacturers, and the FDA, we could really test this process out in a completely different way than with any other type of testing."
But the study confirms a fact of even greater long-term importance to blood centers: "NAT is a very useful platform that we can apply to any emerging agent for which the infectious agent circulates in the blood," she says.
Anne Paxton is a writer in Seattle.