In an era of medicine noted for the development of targeted therapies and personalized health care (PHC), the field of diagnostics has unprecedented opportunities to make a valuable contribution to patient care. It is widely accepted today that a PHC model—enabled by high-value diagnostic tests and molecularly targeted therapies—will be the key to driving further progress in the treatment of many diseases, particularly cancer.

This recognition of the extraordinary potential PHC has to influence patient outcomes is, in turn, causing the value of diagnostic tests to increase across the entire spectrum of patient care, from screening to therapy monitoring. Rather than simply providing information regarding the presence and classification of disease, innovative diagnostic tests are being used to directly inform patient management decisions, such as which targeted therapies to prescribe or whether a patient should be treated more aggressively. Nowhere can this trend—and the critical role of diagnostics—be seen more clearly than in the field of oncology.


Despite the medical urgency associated with its status as the number two cause of death in the United States, cancer currently lags far behind other therapeutic areas in terms of patient response rates to major drugs (Figure 1 below). Similarly, while cancer is comprised of many distinct disease states, the category in general has not seen significant advances in the medical value of diagnostic testing until very recently.

Figure 1

As little as 20 years ago, clinicians were making the majority of diagnoses for many types of cancer on the basis of microscopic inspection of tissue alone. In many cases, that left a significant information gap. It was possible to make general categorizations, such as distinguishing between small-cell and non-small-cell lung cancer cells, but it wasn’t possible to see at the molecular level why cells behave differently from a therapeutic point of view. For example, in non-small-cell lung cancer, two types of cells could look the same but respond very differently to treatment. What pathologists needed was the ability to see inside the genetic structure of the cells to identify what “makes them tick”—the disease pathways they follow and the mutation patterns at the molecular level.

Until recently, the state of the art in diagnostic testing allowed clinicians in many cases to answer the diagnostic question—what is it?—but not the more critical prognostic/predictive question: how will this patient do? For example, will adjuvant chemotherapy improve the outcome for a breast cancer patient after a mastectomy? Now diagnostic tests are starting to answer those questions. A few years ago, a polymerase chain reaction (PCR)-based molecular test was introduced that quantifies the likelihood of disease recurrence in women with early-stage hormone estrogen receptor (ER) positive breast cancer and assesses the likely benefit from certain types of chemotherapy. The value that this kind of test offers in informing medical decisions is unprecedented.

In other therapeutic areas, current biomarker-based diagnostics have clearly provided value in identifying the presence of a disease state, but they lack the sensitivity and/or specificity to enable physicians to make informed therapy decisions. For example, classic markers like prostate-specific antigen (PSA) tests can detect cancerous cell growth but are not specific enough to differentiate benign from malignant cellular proliferations. In addition, PSA does not indicate to the clinician whether a prostate cancer is associated with high or low risk.

In addition to the gap in prognostic and predictive tests, there remains a significant need for improved cancer screening tools. Many of the current in vitro diagnostic tests used for screening (eg, fecal occult blood for colorectal cancer) do not have the ideal sensitivity and specificity characteristics for screening tests. In addition, some screening tests (such as a colonoscopy) are expensive and unpleasant for patients, and thus can result in low compliance or utilization. But the field of oncology diagnostics is in a dynamic transition phase, and emerging developments in screening tests have the potential to facilitate earlier detection and improved patient outcomes.


The landscape of oncology diagnostics is shifting rapidly as revolutionary new technologies are delivering the capability to interrogate cancer cells at the fundamental genomic level (Figure 2). The knowledge that genomic sequencing and other technologies can now provide is effectively changing both the diagnostic and therapeutic components of cancer care. Access to genomic data is giving pathologists the ability to layer prognostic and predictive drug response information on top of conventional diagnostic insights and—even more importantly—is enabling dramatic developments in the application of personalized health care.

Figure 2. Molecular sciences provide new insights into disease pathways and allow the development of new diagnostic approaches and targeted treatments.

In the past decade, the terms “personalized medicine,” “biomarkers,” and “translational medicine” have evolved from hypothetical concepts to reality in drug development, basic and clinical research, and clinical practice. And the growing implementation of personalized health care—the systematic use of insights in patient characteristics, disease biology, and diagnostic tests to tailor medicines and to improve disease management—has become one of the key factors driving the increased value of oncology diagnostics.

The adoption of the PHC paradigm has become pervasive across the pharmaceutical industry, particularly in oncology drug development, and biomarker elements are now commonly incorporated into the development process. Accompanying this trend is an associated acknowledgement that the diagnostic component of the equation is as important as the therapeutic one. For example, companion diagnostics, which are tests used to identify a responsive subpopulation of patients, are now pursued routinely as part of drug development for targeted therapies. As a result, the diagnostic test becomes a critical and necessary companion to the drug, potentially enabling significantly higher rates of response in subsets of patients identified by the presence of certain feature biomarkers.


Accordingly, the value of diagnostics in the era of personalized health care is rising. While diagnostics have historically been considered as commodity reagents and tests, the PHC model is creating a new class of diagnostics that offer medical value by going beyond basic diagnosis to deliver medical information that is used to make specific decisions about patient management.

Figure 3. Identifying the molecular causes of a disease enables researchers to develop targeted diagnostic methods and medicines adapted to the genetic constitution of these degenerate cells.

For example, the results of companion diagnostic tests can indicate directly which specific targeted therapies a patient should or should not receive. Last year, in its first-ever joint approval of a paired drug and diagnostic, the FDA approved a PCR-based in vitro diagnostic test for the BRAF V600E mutation to identify metastatic melanoma patients most likely to benefit from the companion drug that was developed to inhibit that specific mutation pathway. Shortly after that, the FDA approved another drug, an oral anaplastic lymphoma kinase (ALK) inhibitor and the companion diagnostic, a fluorescence in situ hybridization (FISH) test designed to detect rearrangements of the ALK gene in non-small-cell lung cancer. Similar molecular tests by IVD manufacturers are designed to detect epidermal growth factor receptor (EGFR) gene mutations as predictive biomarkers in patients with lung cancer and KRAS gene mutations for select types of colorectal cancer. And HER2 protein tests have been used for several years to identify breast cancer patients whose tumors are most likely to respond to targeted drug therapy.

The PHC-driven revolution in oncology diagnostics, however, goes beyond companion diagnostics. Multianalyte, genomic-based prognostic tests that indicate the risk of disease progression in the absence of therapy, such as the PCR test assessing the likelihood of breast cancer recurrence mentioned earlier, provide tremendous value in informing therapy selection and other patient-management decisions.

Next-generation molecular screening tests promise to help identify diseases and patient risk at earlier stages, when conditions are potentially more amenable to treatment and cure. For example, recent clinical trials have demonstrated the ability of molecular genotyping tests to identify the individual high-risk types of human papillomavirus (HPV)—in particular HPV 16 and 18, which are responsible for about 70% of cervical cancers. The ATHENA study, which involved more than 40,000 women, showed that one in 10 women who tested positive for HPV 16 or 18 actually had cervical pre-cancer, even though they had a normal Pap test. Results like these have had such dramatic influence on the medical community that several leading societies issued new guidelines for cervical cancer screening last year, including the recommendation for HPV genotype co-testing (with a Pap test) for women between the ages of 30 and 65.


A lab technician prepares a next-generation sequencing instrument for a sample run; NGS is transforming cancer discovery research by providing new insights into pathologic signaling pathways.

The influence of recent advances in oncology diagnostics also extends into the realm of therapy decisions. The combination of assays like HE-4 and CA125 (ovarian cancer) in novel algorithms (such as ROMA, or Risk of Ovarian Malignancy Algorithm) can inform physicians on the risk of malignancy for ovarian mass and therefore provide them with an improved ability to decide on an appropriate surgical treatment. And FDA-approved pharmacogenetic biomarkers are being used to identify subpopulations of patients that are at a high risk for drug toxicity based on polymorphisms in cytochrome P450 drug-metabolizing enzymes. Even the diagnosis of disease is being rethought, as traditional anatomical-based classification schemes are evolving into more informative systems based on specific molecular etiologies.


The factor most likely to drive future developments in oncology diagnostics and treatment is the ability to gain greater insight into cancer’s genetic pathways. Few technologies have the potential to influence this direction more than next-generation sequencing (NGS).

During the past few years, NGS has seen widespread growth into expanded applications, in large part due to advancements in technology, particularly in the areas of benchtop sequencing and high-throughput automated systems, resulting in greater ease of use and decreased cost. The area that may be poised to benefit the most from the advances in NGS technology is translational research, the process of translating scientific discoveries into routine applications, including clinical diagnostics. The ability to examine consequential portions of an individual’s genome in order to identify individualized risk predictions and treatment decisions is now within reach and has the potential to impact clinical decision-making globally.

In particular, next-generation sequencing is playing a transformational role in cancer discovery research. Providing new insights into pathologic signaling pathways is one example of advancements made through NGS that were previously impractical due to technological limitations. The advantages of NGS technology are particularly relevant for tumor genetics, where the ability to tease apart the constituent DNAs in a mixed population of tumor cells is especially useful in identifying genetic variants in complex and heterogeneous diseases. In addition, with mutation detection emerging as a key enabler of PHC, diagnostic technologies that provide sensitive, specific, reproducible, and low-cost detection of the full range of mutations present in cancer will be critical for the field to progress.

Despite the tremendous progress in translational research, several significant challenges must be overcome before next-generation sequencing can become a part of routine clinical practice. Manufacturers, laboratorians, and regulatory agencies first need to reach greater consensus in several areas, such as the criteria for selection and quality of samples, data quality and reproducibility, workflow standardization, bioinformatics handling, regulatory guidance, and the clinical significance of variants. In the meantime, research using NGS technology is expected to drive rapidly toward a future that holds greatly enhanced potential for personalized health care. And initiatives to marry gene sequencing with more traditional pathology approaches might offer the best of both worlds.


Another area that represents both challenge and opportunity in the future of oncology diagnostics is the phenomenon of acquired resistance, which is closely related to the development of companion diagnostics. A common tumor response pattern to targeted drug therapy is to show a very dramatic initial response but then stop responding. Because these patients typically recur after a drug’s initial success, this phenomenon has become known as acquired resistance. It appears to occur because the tumor essentially evolves and other cells in the tumor that are not responsive to the targeted therapy, through natural selection and clonal evolution, grow out, and from a clinical perspective that manifests itself as a recurrence.


In addition to microscopic inspection to identify a type of cancer, pathologists are using new molecular tests to help clinicians predict how a patient will do on a certain therapy.

As a medical model, personalized health care will likely have limited success in oncology unless a solution is found to combat acquired resistance. The answer will require both diagnostic and pharmaceutical contributions; the diagnostic to identify the mechanisms responsible for the resistance, and the drug—or more likely the combination therapy—to target the different molecular drivers. So the patient management protocol might involve a paradigm shift from single to multidrug therapy—not a random “trial and error” drug combination, but one that is based on a clear genetic rationale and is tied to the profile that is determined via innovative and intelligent diagnostic tests.


The diagnostics model in oncology will likely need to shift away from a single-test approach as well, to integrate a variety of tools and technologies. For example, pathologists have had the ability for decades to correlate the expression of proteins with histomorphologic patterns by using immunohistochemical methods. Recently, because of the advent of light microscopy-based in situ hybridization methods, gene amplification has been added to their repertoire, allowing them to visualize individual copies of genes and amplification states in the context of morphology. This allows for distinction between different aspects of the specimen, such as invasive and in situ tumor components.

For the first time, pathologists are able to generate genomic profiles of patients’ tumors that incorporate histologic pattern information. While this represents a major advance in the field of pathology, it is only the first of many types of molecular genomic technologies that pathologists will need to acquire. As health care professionals expand their knowledge of the differences between cancer cells and their normal counterparts at the DNA, RNA, and protein levels, it is becoming clear that the characterization of tumors as part of the personalized medicine paradigm will need to occur in a comprehensive manner and will not be possible with single markers or technologies. The evolving picture of personalized medicine for non-small-cell lung cancer is a recent example supporting this notion.

With the potential fragmentation of a diagnostics approach that this development represents, it is also clear that the diagnostics community will need to play a greater role in integrating the delivery, validation, and interpretation of results from different test technologies to make them relevant and actionable for physicians and enhance the medical value they provide. The pathologist, in collaboration with the laboratory staff, can take the lead by helping to facilitate the continued education of physicians, and by extension their patients, regarding the emerging technologies and how they may be used to aid in the detection of disease, stratification of patients, and monitoring of patient response to therapy.


As the personalized health care paradigm redefines how diseases are diagnosed, classified, and treated, the diagnostics field is playing a much more significant role in the practice of medicine. Rather than being independent of therapeutics, diagnostics are now recognized as essential components of the development and prescription of targeted molecular drugs. Of course, several challenges—ranging from technical, quality, and patient safety issues to regulatory and reimbursement obstacles—still need to be overcome before personalized molecular diagnostics achieve their full potential as gateways to patient management. But much progress is being made in these areas, and continued vigilance by the medical community will allow diagnostics to deliver innovative solutions that offer increased medical value and drive optimal clinical outcomes for patients.


Ulrich-Peter Rohr, MD, PhD, is head of Medical and Scientific Affairs, Diagnostics Division, F. Hoffmann-La Roche Ltd, Basel, Switzerland.

Eric Walk, MD, FCAP, is senior vice president, Medical and Scientific Affairs, and chief medical officer, Ventana Medical Systems Inc, a member of the Roche Group, Tucson, Ariz.

Volker Teichgraeber, MD, is senior medical director, Oncology, Medical and Scientific Affairs, Diagnostics Division, F. Hoffman-La Roche Ltd, Basel, Switzerland.

Oliver Liesenfeld, MD, is chief medical officer, Roche Molecular Diagnostics, Pleasanton, Calif.

Richard Batrla, MD, PhD, is global medical leader, Medical and Scientific Affairs, Roche Professional Diagnostics, Rotkreuz, Switzerland.

Johnnie A. Lee, MD, MPH, FACP, is medical director, Medical and Scientific Affairs, Roche Diagnostics Corp, Indianapolis.