Feature Story

Good is getting better in HIV viral load testing

April 2003
Anne Paxton

“Measuring viral load—the amount of HIV-1RNA in plasma—has been an essential part of tailoring HIV combination therapy to individual patients for years. It may seem surprising for a disease once considered hopeless, but the advances in therapy have been so remarkable that diagnostic tests have been pressed to keep pace.

Improving viral load testing is “a major clinical and research priority,” says Angela Caliendo, MD, PhD, director of Emory Medical Laboratories at Atlanta’s Emory University School of Medicine, and associate professor of pathology and laboratory medicine.

“The foundation of all HIV care is the so-called cocktail combination therapy,” notes Tadd Lazarus, MD, director of medical and scientific affairs at Roche Diagnostics, “and it’s all based on knowing patients’ viral load. We would never have known what works in terms of making a cocktail without utilizing that.”

“The drugs are so good, and we’ve become so expert at combining them—making them palatable, reducing the frequency of treatment, and reducing side effects— that more patients are in the realm of either having undetectable levels or are in the very lower limits of the test.”

Roche’s Amplicor HIV-1 Monitor test has long served as the leading test and is highly sensitive. It can measure HIV-1 RNA up to 750,000 copies per mL and is FDA-approved to measure it down to 50 copies per mL with a 95 percent confidence level.

But with the new version 1.5, which received FDA marketing approval last July, the Amplicor promises to “virtually eliminate” falsely elevated or falsely low viral loads in HIVmonitoring, Dr. Lazarus says. Amplicor version 1.5 detects a wider range of HIV subtypes while providing 100 percent specificity.

Dr. Lazarus, an internist specializing in HIV primary care, uses Roche’s technology at one of the largest AIDS centers in the United States, St. Vincent’s Hospital in New York City’s Greenwich Village. He maintains that the key to AIDS treatment is “being able to monitor patients consistently, reproducibly, and accurately at the most sensitive level, and that’s why this test is so accepted and why it’s trusted by pharmaceutical companies. The name of the game is getting people under 50 copies and keeping them there—and making that determination reliably over and over.”

William Valenti, MD, considers the Amplicor 1.5 a “significant achievement” for HIV-1 treatment. “Research has shown us the importance of measuring viral load levels to as low as possible to improve outcomes for patients,” he notes, adding that detecting more subtypes has also become increasingly important.

Dr. Valenti, founding medical director of Community Health Network and clinical associate professor of medicine at the University of Rochester School of Medicine and Dentistry, has been using both versions of the Amplicor HIV-1 Monitor test for more than five years. He describes Amplicor as a dynamic technology in part because it tests viral load rather than T-cells. “T-cell counts don’t shift as quickly as viral load,” he points out, and viral load will increase fairly rapidly if the patient is off-therapy.

Class B subtypes of HIV-1 still predominate in the United States. Ninety-five percent of the subtypes here have been subtype B, Dr. Valenti says, but every year there is a small increase in non-B subtypes. “There’s enough of an increase to take another look at the technology we’re using to monitor viral load.”

Amplicor version 1.5 can detect HIV group M subtypes A through H as well as B. “We realized that we needed to modify the test so there is equal amplification across all the major subtypes,” Dr. Lazarus explains. “We also felt it was important to have our HIV product on a Cobas amplification platform.”

Automation is far removed from the state of the art in HIV management as it existed 15 years ago. When Dr. Caliendo began working on HIV treatment, “there were few antiretroviral drugs, no viral load testing, and no resistance testing,” she recalls.

Dr. Lazarus, too, remembers those years. “There are some patients I’ve been treating for a very long time, and it’s very gratifying to see they’ve made it through the era of darkness and desperation.” In the days of monotherapy or dual therapy, laboratory diagnostics functioned mainly to measure the damage to the immune system. “We were only measuring the T-helper cells, watching them decrease, as the only indication of treatment failure,” says Dr. Lazarus.

Now, he says, we are in an era where HIV viral load testing is a crystal ball to look inside the patient, and to know and respond to the patient before the immune system is damaged.

Can the new version of Amplicor lower the cost of treatment? “We’re in the middle of clinical studies to document savings,” says John McCune, Roche’s marketing manager for HIV molecular diagnostics. “But the reality is that extra cost comes in when your test is not 100 percent specific. You get false elevations, and patients are upset at seeing elevated viral loads. Then you have the cost of extra testing and potentially unnecessary changes in therapy.”

Dr. Lazarus also stresses the cost-effectiveness of highly accurate viral load testing. “We have and must continue to strive to find ways to help patients adhere to often complicated treatment regimens, and one way I’ve always found to be extremely successful is patients’ viral load. It’s an extraordinary reward system emotionally for the patients to continually see their levels be undetectable or barely detectable. So coming in every three months to see that gives people at very little cost the motivation to keep going.” Equally important, he adds, is that “it acts as an early warning, where a slight perturbation in viral load can let a physician intervene and find out what’s happening in the patient’s life.”

Almost always, such mild perturbations are caused by the patient being a little less compliant with the therapy. “Of course I can’t say they might not also represent early changes in resistance patterns,” Dr. Lazarus says, “but the two go hand in hand, because proper selection of antiretroviral drugs with almost 100 percent compliance means you stave off resistance as long as possible.”

Not every patient benefits from the current approaches to AIDS treatment, Dr. Lazarus notes, especially those who have been infected with HIV for a long time. These “highly experienced” patients may have lived through the sequential monotherapy age, after which AZT (now called zidovudine) came on the scene, then DDI (now called didanosine). “They would sequentially take every new drug in the monotherapy era before people knew about combining drugs. Now we know we must combine drugs from different classes, and usually we have at least three drugs on board,” Dr. Lazarus says.

Roche has been committed to making viral load testing available locally, he says, since many smaller and regional hospitals now send their tests to a reference laboratory. “One reason we worked on having an automated platform was that we recognized it meant these hospitals could be doing HIV viral load testing locally as part of their own commitment to caring for AIDS patients,” Dr. Lazarus says. Soon the company will have front-end automation, so the specimen will go in, the nucleic acid will be extracted, and it will then be tested by Cobas Amplicor.

Whether the facility is a hospital-related AIDS program, a clinic, or a private practice, it always has to determine what to do internally and what to send out. “But in my five years in this position,” says Dr. Lazarus, “I’ve seen that what people called ‘esoteric’ five years ago in many places is now routine. So a molecular revolution has with good speed penetrated local and regional facilities, because they recognize that molecular testing methods provide the future basis of care—whether in infectious diseases and being able to provide same-day molecular-based methods to determine strep throat in a child, or moving toward rapid and accessible markers for oncology.”

Patients’ degree of resistance is a legacy of the inadequate therapies of early AIDS treatment, he notes. There were inadequate combinations of drugs, especially in the pre-HIV viral load era. But when protease inhibitors came out just as HIV viral load testing became the basis of the cocktail, they were still all in the same class of drugs.

Roche developed the first drug in what was a new class of drugs—protease inhibitors. That drug, saquinavir, when combined with the original class of HIV antiretrovirals, nucleoside reverse transcriptase inhibitors, formed the cocktail that has revolutionized HIV and AIDS treatment. “Before that time, just taking one or two drugs was ineffective and it led to massive mutations and resistance. So patients who bear the genetic fingerprint of resistance and patients who have gone through periods of imperfect adherence both have potentially high levels of resistance,” Dr. Lazarus says. He calls resistance a “potent issue in terms of the road that must still be traveled” to make living with HIV or AIDS no different than living with more benign chronic conditions.

The precision of Roche’s Amplicor versions 1.0 and 1.5 is similar, as are other performance characteristics, though there is a slight increase in the linear range of the ultrasensitive assay, from 75,000 with version 1.0 to 100,000 with version 1.5, Dr. Caliendo says. “The major advance of 1.5 is quantifying the non-B subtypes. That’s why the assay was designed.”

Detecting non-B subtypes is not a major issue in the United States and Europe. However, Dr. Caliendo says, “we are seeing more patients with non-B subtypes. And it is becoming an issue globally.”

Two other assays with different technology are also FDA-approved and available: Bayer’s bDNA test and BioMérieux’s NucliSens HIV-1 QT quantitative viral load assay, which uses the firm’s proprietary Nucleic Acid Sequence-Based Amplification, or NASBA, technology.

Users like Gregory J. Tsongalis, PhD, director of molecular pathology at Hartford Hospital and associate professor at the University of Connecticut, point to some drawbacks of Amplicor. Dr. Tsongalis’ laboratory switched last summer to the Bayer test for reasons of logistics and cost.

“Like most people, we were using the 1.0 manual Roche assay and we were only performing the standard assay, not the ultrasensitive, because logistically it would have been a nightmare for us,” he explains. “You’re at the mercy of physicians to decide which test they want. Some laboratories have set up algorithms based on the patient’s last results, but it just takes way too much time for the technologist to do.”

“The other issue we had was if you had to run one test and reflex to the ultrasensitive, or vice versa, then payers would only pay for one test, and the laboratory would have to eat the extra cost. The reason is you get the same patient specified, but there is only one CPT code for viral load testing.”

The bDNA test had its pluses and minuses, he says, including a slower turnaround time, and his laboratory waited through several versions before switching. When the bDNA version 3.0 came out in 2001, however, “that version addressed all our concerns. It was one test that covered the dynamic range from ultrasensitive up to the higher end, and the format of the assay gave us higher throughput, so we could run more in one run than with the Roche assay.” With Amplicor, his laboratory was running 20 samples per run, three to four times per week, but with bDNA, it could run 84 samples per run, only once per week. Per patient sample, he adds, the bDNA assay was about $30 less. “For us that’s a substantial savings,” he says.

The main concern of physicians when the laboratory told them about switching the test was how to interpret the bDNA results—but the hospital resolved that, Dr. Tsongalis says, by adopting the same reporting format.

To simplify logistics, Dr. Tsongalis would like to see turnkey viral load testing technologies. “We need to be able to put the samples on the instrument, walk away, and come back in an hour,” he says. Right now both Bayer’s test and Roche’s Amplicor test take his laboratory about a day and a half to perform. Dr. Caliendo’s laboratory at Emory uses the Amplicor test, and testing there is completed in a single day. But the turnaround time is usually determined, she says, not by how long it takes to do the test but by whether the laboratory has a large enough volume to batch the tests frequently. “Some clinicians want their results within the week, while some may wait 10 days to two weeks,” she says.

Roche’s McCune says Amplicor is the only test to measure 50 copies per mL at the low end with 100 percent specificity. Michael Cronin, BioMérieux’s clinical marketing manager, says NucliSens is a target amplification assay like Amplicor and also has 100 percent specificity. While NucliSens is not perfect in detecting non-B subtypes, he says, it will cover the entire dynamic range—50 to more than 5 million copies per mL—with “one sample, one test.”

But, says Dr. Caliendo, “if you judge the sensitivity on the ability to detect the virus in 95 percent of replicates, the sensitivity of NucliSens is more in the range of 175 copies per mL.”

Though she believes both tests are well designed, she points to carryover contamination as a potential problem for the Amplicor and NucliSens assays. “But the assays have been designed to minimize the problem,” she adds.

Bayer’s bDNA detects subtypes A through G, and studies have shown that NucliSens detects most subtypes, but “there seems to be some data pointing to underquantification of subtype G,” Dr. Caliendo says.

Performance specificity also varies. bDNA’s lower specificity is due to assay chemistry, she says. “You can occasionally get a result of viral load in a specimen that really is negative, and these are usually low-level positive specimens, with less than 1,000 to 2,000 copies per mL.”

As with most tests, volume and turnaround time are key drivers of cost for all three assays. McCune believes their cost is essentially the same—an average of $60 per test, depending on how many a lab is performing and what other molecular assays it is using. Batch size and physician-requested turnaround time, among other things, affect cost.

Getting on a real-time PCR platform will be an important advance for the Roche test, Dr. Caliendo says. “That would give us a broad dynamic range with a single test,” thus making the dynamic range equivalent to those of the other two tests, she says. Roche expects clinical trials to begin in July on its real-time PCR TaqMan assay.

Outside the United States, BioMérieux is selling a new version of NucliSens based on real-time NASBA with molecular beacons, Cronin says. A molecular beacon is a modified probe put into the NASBA reaction. When the RNA is amplified, the molecular beacon probe will anneal, or bind to, the amplified RNA, generating a signal at the same time. “That’s where the term ‘real-time’ comes from,” he says.

For now, with three FDA-approved tests for viral load, laboratories have good alternatives to choose from. In the future, Dr. Caliendo says, laboratory professionals can look forward to more advances in viral load testing. “Expanding the dynamic range and automating the tests should be where the focus is now,” she says.

Anne Paxton is a writer in Seattle.