Every six months, the software programmers at the molecular-diagnostics business unit of Bayer Diagnostics go back to the drawing board. The programmers responsible for modifying the treatment algorithms for the company's HIV-1-resistance software must update their algorithms regularly because the virus mutates — and acquires drug resistance — so quickly.
The programmers are just one of the groups of professionals Bayer employs to keep its tests up to date. HIV clinicians and researchers regularly meet to share scientific data, dissect high-profile papers and define the impact of new mutations on the virus's resistance to current therapies. At the end of their meetings, the programmers receive a new set of rules, which they translate into new algorithms, transforming the molecular analysis of the virus and its treatment resistance into clinical information that is useful to doctors caring for patients. By the end of the journey, the molecular data have come full circle, back into the hands of the doctors on the frontline of treatment.
Like genomics and functional genomics, molecular diagnostics relies on a network of scientists to make sense of this data-rich field. In addition to programmers, companies well-positioned to make an impact in the molecular-diagnostics market are employing molecular biologists, chemists, biostatisticians and even engineers to bring a test from the research and development (R&D) phase and release it into this fast-growing market.
Detection and Prediction
In 1985, the US Food and Drug Administration (FDA) approved a nucleic acid-based diagnostic test for legionnaire's disease from San Diego-based Gene-Probe. Most of the tests that followed in the 1980s and early 1990s targeted infectious diseases, and tests for HIV, hepatitis C and human papilloma virus (HPV) remain in strong demand today. Lately, molecular tests that diagnose and predict cancers have become more profitable targets. Cardiovascular disease has become another popular target as rates of heart disease increase.
The HPV test marked a shift in the definition of diagnostics, because it not only identified the infection, but predicted a woman's risk of developing cervical cancer. Now the field of molecular diagnostic has developed to include prognostics and biomarkers. The shift from simple detection to prediction has broadened companies' needs. Companies designing prognostic tests that use DNA microarrays may have a greater need for engineers, whereas those that sift through massive amounts of data linking genotype to treatment outcome require biostatisticians and researchers with clinical backgrounds.
Growing Fast
AFFYMETRIX
Affymetrix is collaborating with Roche to develop its microarrays for diagnostic detection of single-nucleotide polymorphisms.
Historically, the diagnostics industry has been overshadowed by the excitement surrounding new drugs. But times are changing, says Joffre Baker, chief scientific officer at Genomic Health in Redwood City, California. "There's an enormous nascent industry here," he says. "Genomics, sequencing and technologies allowing functional genomics have changed everything."
The US market for molecular diagnostics generated revenues of $1.3 billion in 2000 and was predicted to generate about $4.2 billion by 2007 (see Nature Biotechnol. 20, 6–7; 2002). And this means employment. Roche Diagnostics, based in Basel, Switzerland, expects to recruit 250–300 graduates over the next year. Roche is not alone in diagnostic growth. In 1998, Bayer acquired the molecular division of Chiron diagnostics and began its foray into molecular diagnostics. Since then, Bayer's molecular diagnostic tests for HIV and hepatitis C have secured a major share of the market.
Peter Knueppel, head of molecular diagnostics at Bayer Diagnostics in Tarrytown, New York, says that although Bayer's molecular-diagnostics business is not currently looking urgently for staff, the company always has jobs for strong candidates. Molecular biologists are in highest demand, but chemists and software engineers are an important part of the mix. Apart from education, he says, "you need to have a high degree of flexibility and be willing to move quickly — it's a fast-moving industry".
Currently, a shift in disease targets is causing much of the movement within the industry. With plenty of molecular tests available for infectious diseases, companies are moving on to oncology and cardiovascular-disease diagnostics. In 2001, Abbott Laboratories of Abbott Park, Illinois, acquired the diagnostic company Vysis and began to expand into the oncology arena. In 2002, Abbott began a partnership with Celera Diagnostics of Alameda, California, which is likely to help it secure a larger slice of the molecular-diagnostics business (see 'Strong alliances').
Talent Search
"We're looking for PhDs with industry experience in molecular biology, chemistry and engineering," says Ed Michael, president of Abbott's molecular-diagnostic business unit. Michael says that the molecular-diagnostics division has 40 positions open now, ranging from research to marketing. The company's clinical-research group needs statisticians for number crunching and analysis, and graduates with industry experience could find themselves walking into a senior position at one of Abbott's two campuses near Chicago.
Researchers at Genomic Health are betting their future on molecular tests for cancer. The company, which was spun off from the genomic-information business Incyte, launched its first commercial diagnostic test in February. The 21-gene test predicts the risk of recurrence of certain types of breast cancer. "The test helps identify patients who have a low risk of recurrence and keep them from being exposed to unnecessary chemotherapy," says Baker.
In addition to diagnostics and prognostics, molecular tests are being used to discover patients' responses to treatment. Roche is among the leading companies moving beyond the one-size-fits-all approach to medicine. Last June, in a partnership with Affymetrix of Santa Clara, California, the company launched a DNA chip-based assay test that will identify polymorphisms in two of the drug-metabolizing genes, CYP2D6 and CYP2C19. The test was designed to help doctors choose the best medicine and dosage for their patients, to improve treatment and avoid some of the 2 million adverse drug reactions that occur in the United States every year and the 100,000 deaths they cause.
The value of genomics may be realized much faster in diagnostics than in therapeutics, but it is still not an easy feat. "Some of the publications coming out of laboratories are nice, but the clinical relevance is sometimes questionable," says Janet Warrington, vice-president of clinical and applied genomics R&D at Affymetrix. "They may be discovering unique biology or pathway information, but it is not an easy step to go from there to a new test." Cost is also an issue. Molecular-diagnostics tests are expensive and in the United States many are not covered by healthcare insurance. But in the long term, says Baker, the predictive nature of these tests could have a huge economic impact, as they could lead to earlier, less expensive treatments.
As the field grows, companies are employing recent graduates in positions from research to manufacturing to marketing. Industrial or clinical experience is usually essential, but collaboration with other sciences is a priority. "The ideal job candidates are people who have a lifelong interest in this area and who have the ability to pick up information that is outside your primary discipline," says Warrington. Biochemistry, genetics and chemistry graduates predominate in her research group, she says. "They must have some facility using current analytical packages and be able to communicate."
Although molecular biologists predominate in the industry, engineers are equally important, says Schewe. "People don't readily imagine us as an organization recruiting engineers, but it's logical. The instrumentation can be very challenging." Roche's global organization also requires people with worldly skills — including language proficiency and cultural sensitivity. "It's not the scientific expertise alone any more that makes things work. Success or failure depends on how people work together," says Schewe.
