Improving health care and laboratory medicine: the past, present, and future of molecular diagnostics

In vitro diagnostics has many different disciplines. Histology tests are classic for diagnosis, followed by chemistry and serology tests. The newest in vitro diagnostic tools are molecular. These clinical tests are used to determine therapy type, duration, and dose; they are also used to determine if therapy has been successful. The choice of test depends upon the question being asked.

As molecular diagnostics has evolved, it has demonstrated clear advantages over some traditional methods, although it does not completely replace other methods. Molecular methods are extremely sensitive, which is critical for the direct detection of viral nucleic acids. When a pathologist states that patients have undetectable virus, that does not mean that they have no virus in their systems. It means that the level of virus is below the limits of detection for the assay used. A robust and sensitive assay can provide more useful information. Other advantages are that molecular assays often require minimal sample volumes and do not require culture. Molecular methods can be very accurate, as they measure DNA or RNA sequences specific for the virus or mutation of interest. Finally, molecular diagnostics usually offers tremendous time savings in the laboratory, allowing clinicians earlier access to data and thus earlier treatment for patients.

Nucleic acid amplification is the key to molecular diagnostics. A molecular test for a virus must detect a specific RNA or DNA sequence in the midst of a tremendous amount of other nucleic acids: human genomic DNA, mRNA, bacterial DNA, and RNA. This process is like pulling the proverbial needle out of the haystack. The power of a nucleic acid amplification technique is that it can amplify a specific sequence and do so in a predictable manner so that the number of molecules in the initial sample can be quantified.

Polymerase chain reaction (PCR) is the gold standard for amplification processes in diagnostics. Since the technique was first published in 1985 (1), it has become the most widely used nucleic acid amplification technology. With PCR, a target sequence of nucleic acid is amplified through a repetitive series of reactions catalyzed by a single enzyme, a thermal stable nucleic acid polymerase. PCR can amplify both DNA and RNA.

The process of PCR typically involves taking the sample through three different temperatures, each of which causes a different event. The sample is placed into a tube with the proper primers, thermal stable polymerase, nucleotides, and buffers. In the first step, denaturation by heating to 95°C causes the double-stranded DNA to form single strands. In the second step, the temperature is dropped to 55°C to 60°C, which allows the target specific primer to bind to only the intended nucleic acid sequence. The specificity of an assay depends on the quality of primer choice. Finally, in a third step, the temperature of the reaction is raised to about 72°C and the polymerase synthesizes an identical copy of the target sequence, called an amplicon. The cycle is repeated many times through the three different temperatures. Everything occurs rapidly in a closed tube in a thermocycling instrument.

PCR causes exponential amplification of a target nucleic acid region. Starting with 1 copy of a molecule leads to 16 amplicons after 4 cycles. Starting with 2 copies leads to 32–4 copies after the first cycle, 8 after the second cycle, 16 after the third cycle, and finally 32 after the fourth. Clinicians do not know how many copies are present in a sample but use the result to extrapolate back. If they found 96 amplicons after 4 cycles, they could de-termine that the sample had 6 copies initially. Instrumentation now allows this to be done automatically.

Roche has developed a portfolio of PCR business units based on disease areas. In six areas, an expertise is already present: virology, women's health, genomics, microbiology, blood screening, and oncology. This article discusses applications of molecular diagnostics in these fields.

VIROLOGY: HIV AND HEPATITIS

The effectiveness of HIV drugs has been determined by molecular methods, which measure the viral loads. The PCR-based Roche Amplicor (1992) and COBAS Amplicor (1995) HIV-1 Monitor tests were instrumental in the Food and Drug Administration (FDA) approval of most HIV drugs on the market. In March 2004, the National Institutes of Health released revised Guidelines for the Use of Antiretroviral Agents in HIV-1 Infected Adults and Adolescents (2). The document stated: “The goal of therapy is suppression of viral replication to below the level of detection.” Whereas the 2002 guidelines said that physicians should use the most sensitive assay available, the 2004 guidelines stated: “Results of therapy are evaluated through plasma HIV RNA levels.” These statements highlight the clinical utility and importance of diagnostics in directing and monitoring therapy.

Because of its increased sensitivity, the Amplicor HIV-1 Monitor assay can reveal earlier when the antiretroviral therapy is failing,because viral loads that would have been below the limits of detectability with another assay are detectable with PCR. Such knowledge helps the clinician intervene earlier and keep the patient healthy.

Dr. Essmyer from St. Luke's Medical Center in Kansas City shared her data with me. In 1996, before HIV viral load testing was used to direct therapy,the patient pool had widely variable viral loads, and therapy was not controlling viral replication. Between 1997 and 2001, physicians began routine use of viral load testing to monitor treatment and adjust antiretroviral regimens. With such informed treatment, almost all patients had nondetectable viral loads (Figure 1). Those patients who continued to have a high viral load (> 100, 000)appear to have drug-resistant virus.

Figure 1.

Viral load (a) in 1996, before molecular viral load testing, and (b) from 1997 to 2001, when molecular viral load testing was routinely used. With the molecular data, clinicians were able to better control viral loads of their patients with HIV by adjusting therapy. Data courtesy of Dr. Essmyer, St. Luke's Medical Center, Kansas City.

Hepatitis C virus (HCV) is another growing global health problem. Estimates suggest that 170 million to 200 million people worldwide are carriers. HCV is four times more prevalent than HIV. As many as two thirds of individuals infected with HCV have no symptoms and are therefore unaware of their infection. Unfortunately, symptoms usually appear decades after the virus was acquired. Over 280, 000 deaths annually are associated with the effects of HCV infection, such as cirrhosis, fibrosis of the liver, and hepatocellular cancer (3–6).

There are three major types of molecular testing: qualitative, quantitative, and genotyping. Each helps answer a different clinical question. A highly sensitive HCV qualitative assay can determine if HCV is present and active a quantitative viral load test can determine how much HCV is present and a genotype assay can identify the HCV type. Diagnostics is quickly changing with the introduction of real-time PCR and the development of new clinical tests.

A treatment algorithm is used by many physicians for patients with genotype 1 HCV (Figure 2). Molecular diagnostic tests are critical in determining treatment course and duration. These tests help physicians identify patients who are unlikely to respond to therapy in the long term and justify the discontinuation of therapy. The process begins with the serological determination of HCV exposure and a molecular test to detect the presence of HCV RNA, which indicates an active infection.Since newer real-time PCR quantitative molecular tests have excellent sensitivity, some laboratories now use quantitative tests to confirm viremia. Then, if genotyping reveals HCV genotype 1, the original quantitative test result can be used as the baseline viral load before treatment. If a qualitative test is used to confirm viremia, a quantitative test should be performed before treatment to determine dosing of ribavirin. After 12 weeks of treatment with pegylated interferon and ribavirin, another quantitative assay is performed. If the patient 's HCV RNA levels do not drop by two logs, therapy may be stopped, since the chance of achieving a sustained viral response with further treatment is very low. If a two-log drop is achieved, another 12 weeks of therapy should bring the HCV RNA levels to below detectable limits. A complete round of therapy —48 weeks—will often achieve a sustained viral response in patients who attain the 12 and 24-week milestones. In monoinfected genotype-1 HCV patients, 40% to 50% of patients have been reported to have a sustained viral response (7).

Figure 2.

Treatment algorithm for hepatitis C genotype 1, targeting treatment with the use of molecular diagnostics. Based on reference 8.

Use of the algorithm can result in cost savings and decreased morbidity. The retail cost of 48 weeks of treatment of pegylated interferon and ribavirin is $29, 000. Discontinuing therapy after 12 weeks in the group that has been shown not to benefit translates to a savings of $21, 750 per patient. When therapy is discontinued at 24 weeks, the savings is $14,500. Further, the therapy has significant side effects; because of the morbidity, patient compliance is often low. It is easier to motivate patients to try the treatment for the shorter time periods when they know that its effectiveness will be evaluated. Similarly, those in whom the treatment is effective become motivated by the test result since continued treatment may lead to a sustained viral response.

A good example of the failure of serological tests alone is shown by a case study provided by Dr. Wohlfeiler from Florida. A 51-year-old man had been HIV positive for more than 15 years. He lived with an HIV+, HCV+ partner for 18 years. The patient was taking several antiretroviral drugs and had a CD4 count of 329 cells/mm³ (22.5%). His alanine aminotransferase was 304 IU/L, and his aspartate aminotransferase was 158 IU/L. Hepatitis panels showed the presence of hepatitis B and hepatitis A, but not HCV. The physician ordered a serum test of HCV antibody and a quantitative HCV RNA test. When the former test result came back as negative, the laboratory decided not to run the HCV RNA test. Dr. Wohlfeiler persisted in his request, and the result of the HCV RNA assay was amazingly high: a viral load of > 850,000 IU/mL. Apparently, the HIV infection had affected the patient's immune response to HCV, explaining why antibodies were not detected. Without the molecular test, this patient would have been undiagnosed with active HCV infection.

WOMEN'S HEALTH:CHLAMYDIA

Roche has a PCR test that detects Chlamydia trachomatis and Neisseria gonorrhoeae at the same time from the same sample. A number of regulatory bodies and associations have recommended widespread screening for these diseases (8). For example, the Health Plan Employer Data and Information Set and the Centers for Disease Control and Prevention recommend screening all sexually active women up to age 25 for chlamydia and screening older women if risk factors are present (9).

The screening guidelines reflect the occurrence of the disease. For biologic reasons, women in their teens and twenties are particularly susceptible to this infection (Figure 3). Nationwide screening of women at a family planning clinic showed that in 5 states and territories, the prevalence rate was < 4.0%; in 11, it was 4.0% to 4.9%; and in 37, it was > 5.0%. Texas' rate in this study was 8.2%(10).

Figure 3.

Age-specific and sex-specific rates of chlamydial infection in the USA, 2003. Source: Division of STD Prevention, Centers for Disease Control and Prevention.

Rates of chlamydial infections rose significantly between 1987 and 2001. Several factors played a role in this rise, including increased awareness of the need for screening in public and private health care settings, improvement in the sensitivity of diagnostic tests, and improved surveillance and reporting systems. Screening does make a difference. A large-scale screening program initiated in family planning clinics of region 10 of the Department of Health and Human Services (i.e., Alaska, Idaho, Oregon, and Washington) has been followed by a 60% reduction in chlamydial infection rates. Other screening programs have been initiated based on such demonstration projects (11).

While molecular diagnostic tests are available for Chlamydia trachomatis and Neisseria gonorrhoeae, they have to be compared with other diagnostic methods. Table 1 lists advantages and disadvantages of each method.

Table 1.

Comparison of diagnostic tests for chlamydia and gonorrhea

Test	Advantages	Disadvantages
Culture	• Specificity nears 100%, thereby reducing the potential for false- positive results	• Requires a skilled laboratorian; is labor intensive and expensive
		• Sensitivity is about 80%
		• Cervical specimens only
DNA probe	• More stable transport of specimens	• Cervical specimens only
		• Sensitivity is about 65%
	• Less expensive than culture
Enzyme-linked immunosorbent assay	• Less technically demanding than culture	• Cervical specimens only
		• Sensitivity is about 60%
	• Less expensive
Nucleic acid amplification	• ≥90%sensitivity and specificity	• More expensive than DNA probe or enzyme-linked immunosorbent assay
	• Can use either urine or cervical swabs as specimens

Source: Health Plan Employer Data and Information Set.

PERSONALIZED MEDICINE: PHARMACOGENETICS

The goal of personalized medicine has been touted in the popular press: both Wall Street Journal and US News & World Report have reported on it. Currently, drug therapy often relies on trial and error. Even “state-of-the-art” pharmaceutical therapy is applied in a one-size-fits-all manner. The response rate is far from perfect. With personalized medicine, therapy will be selected based on individual patient characteristics that become known through bioinformatics. The results will be response rates that approach 100%, as well as increased survival rates, improved quality of life, cost savings, and reduced morbidity and mortality.

In October 2000, Fortune published a story of a boy who died because personalized medicine was not available:

The death of nine-year-old Michael Adams-Conroy didn't seem at first like a signal event in medicine…. While recuperating from what seemed to be flu, Michael went into a prolonged grand mal seizure and died. His grieving parents, Jayne and Neil, soon got another shock: An autopsy showed a massive overdose of Prozac in Michael's blood and tissues, raising the specter of a murder charge against them.… Thus began the Adams-Conroys”, painful pilgrimage to a medical frontier known as pharmacogenetics, the study of how genetic idiosyncrasies influence responses to drugs (12).

Michael's Prozac levels were accumulating after every dose; little drug was being cleared from his system. He was a poor metabolizer of drugs through his CYP450 2D6 genotype. Drugs metabolized through the CYP450 2C19 gene make up a quarter of all prescription drugs (Table 2). In contrast to Michael's poor metabolism, some people are ultra-rapid, extensive, or intermediate metabolizers of these drugs (Figure 4) Knowledge of the category of drug metabolism phenotype helps physicians tailor dosing.

Table 2.

Common CYP450 2D6 and 2C19 drugs, which patients metabolize differently

CYP2D6	CYP2C19
Antiarrhythmics	Amitriptyline (in part)
Antidepressants	Certain barbiturates
Beta-blockers	Proguanil
Neuroleptics	Citalopram
Others	Cyclophosphamide
—Atomoxetine	Diazepam
—Codeine	Imipramine
—Ondansetron	Mephenytoin
—Tamoxifen	Omeprazole

Figure 4.

Genotype can have a significant impact on drug metabolism and response.

Information on drug responses can be gained through a molecular method called microarray genotyping analysis using the AmpliChip P450 test, which is currently available as an FDA-approved product from Roche.

MICROBIOLOGY, BLOOD SCREENING, AND ONCOLOGY

As the USA was being afflicted with the scare of anthrax as a bioterrorism agent, a molecular assay was developed and put into use in 4 weeks. That speed is nothing short of revolutionary. Patterns of anthrax infection differ globally as well as geographically within the USA; an assay can be customized to a local or regional pattern.

When the USA faced the West Nile virus epidemic, with a corresponding threat for the blood supply, the FDA and the Centers for Disease Control and Prevention turned to molecular diagnostic companies to rapidly design and manufacture an assay. An assay with an automated platform was made available in only 8 months. Such an accomplishment would have been impossible without molecular methodologies.

One example of the impact of molecular diagnostics in oncology was featured in the New York Times (13). Two types of childhood medulloblastoma have been identified based on clinical course; however, the two types look similar under a microscope. When samples were taken from both groups of patients and put into a DNA microarray system, two different patterns emerged that now allow physicians to provide the right level and type of chemotherapy for these young patients. It spares some children unnecessary morbidity and offers a better chance of a cure for children who can benefit.

FUTURE OF CLINICAL MOLECULAR TECHNOLOGY

Only 51 years ago, Watson and Crick proposed the structure of DNA and alluded to the significance of their findings: “It has not escaped our notice that the specific pairing we have postulated immediately suggests a possible copying mechanism for the genetic material” (14). In 2001, the human genome sequence was published and created an opportunity for revolutionizing health care with individualized medicine, the curing of diseases, and new horizons for preventive medicine.

Real-time PCR is the next evolution of PCR; it is currently in clinical use. As PCR generates specific amplicon molecules, real-time PCR uses a fluorescent reporter signal to measure the amount of amplicon as it is generated (Figure 5). This kinetic PCR allows for data collection after each cycle of PCR instead of only at the end of the 20 to 40 cycles. Real time PCR has an increased dynamic range compared with running the PCR completely and then measuring the amount of amplicon at the end of the reaction.

Figure 5.

Real-time PCR measures at the base of the exponential phase (rectangles) rather than at the endpoint (ovals). The exponential phase is the start of predictable amplification; the earlier the exponential phase begins, the more copies of the product present in the initial sample. After a certain point, a plateau effect occurs. Each line in the table represents a different sample. A negative sample shows up as a horizontal line at the threshold level.

Several tasks lie ahead in molecular technology. First, the genetic correlations with disease need to be validated. They are based on retrospective studies but will need to be based on prospective studies. Second, clear guidance is needed from the FDA on genetic tests. Third, education is needed—for physicians, health care workers, and the general public. It will also be important to ensure the protection of human subjects and patients in this process and to examine the economic value of molecular diagnostics.

We are in the midst of a change in diagnostic paradigm. Now tests are usually used for diagnosis, therapy, and monitoring of therapy. The focus in the near future will be on wellness: predisposition testing, targeted monitoring, and prevention of diseases through nutrition, lifestyle, and medications. The goal is to give people a greater opportunity to attack diseases before they become patients.

Clearly, health care will be more complicated in the future. The complexity will require cooperation among disciplines. A new, shared language will develop based on the molecular foundations of disease. With such knowledge, health care will also become more individualized. As Sir William Osler said, “If it were not for the great variability among individuals, Medicine might be a Science, not an Art” (15).