BB&T Contributing Editor

ORLANDO, Florida — The annual meeting of the American Society of Clinical Oncology (ASCO, Alexandria, Virginia) is the leading venue worldwide for presentation of new developments in cancer management, including new targeted therapies, advances in cancer diagnosis and treatment monitoring, and new approaches for cancer screening and prevention. The theme of this year's conference, held here in late May, was personalized cancer care, reflecting a broader trend in healthcare toward personalized medicine.

Personalized cancer care has emerged only within the past few years, enabled by the development of genomics-based technologies to characterize a patient's cancer on a molecular basis and select therapies which selectively target the disease, improving the effectiveness of therapy while reducing damage to normal tissues and organs. Numerous companies are now providing genomic testing services for cancer management, analyzing large panels of genetic markers numbering over one thousand in some cases, which provide prognostic information as well as data for guidance of targeted therapy.

Progress was reported at the ASCO conference in optimizing the use of conventional cancer therapies such as chemotherapy and radiation to enhance efficacy while minimizing adverse side effects, and in enhancing the delivery of anti-cancer agents using new biological modifiers. New developments in cancer vaccines were announced at the conference, along with advances in cancer imaging, cancer screening, and early detection.

In 2005, cancer was the second-leading cause of death in the U.S., and in 2009 an estimated 1,479,350 new cases of cancer will be diagnosed according to the American Cancer Society (Atlanta). Worldwide, as shown in Table 1, more than 12 million new cases of cancer were diagnosed in 2007. Of those, almost half (46%) occurred in Asia, 22% in the Americas, and 25% in Europe.

Lung cancer is the highest-incidence cancer type worldwide, followed by breast, colorectal, stomach, prostate and cervical cancer. The aging of the populations in developed countries and improvements in life expectancy in developing regions are factors that will drive a continued increase in cancer incidence and prevalence worldwide.

A study published online in the Journal of Clinical Oncology by Benjamin Smith, et al., estimates that cancer incidence will increase by 45% in the U.S. between 2010 and 2030, from 1.6 million to 2.3 million. That trend will create a growing demand for devices and drugs used in cancer management.

The trend for development of targeted therapies and companion tests to select the best targeted treatment for each patient is now well-established. Essentially all new cancer drugs now under development are targeted agents which interdict specific biochemical pathways known to be associated with cancer development, and which rely upon molecular tests to guide selection of responsive patients.

The new paradigm in cancer management has spawned numerous new companies focused on developing genomic cancer tests, as well as companies developing targeted drugs.

Personalized therapy for all patients

At present, only a relatively small percentage of cancer patients are treated using guidance data from genomic cancer tests, since such tests have only become available for clinical use within the past few years. As shown in Table 2, there are now many providers of molecular testing services for cancer management, and certain tests, including the Oncotype DX and the Femtelle uPA/PAI-1 for guidance of adjuvant chemotherapy in breast cancer, are now recommended in the physician guidelines published by ASCO and other professional organizations.

Just prior to the ASCO conference, Genomic Health (Redwood City, California), the provider of the Oncotype DX breast cancer test, reported passing the 100,000-patient milestone, representing more than 7,500 physicians who have ordered the test. Test volume in 2008 for Genomic Health was 39,600, up 62% from 24,450 in 2007.

Since there were roughly 185,000 new cases of breast cancer in the U.S. in 2008 according to statistics from the American Cancer Society, and substantially all of the tests performed by Genomic Health in 2008 were ordered by U.S. physicians, about 20% of new breast cancer patients were tested for their genomic profile using the Oncotype DX assay in 2008.

The Oncotype DX test is not applicable to all patients, since it is only recommended for use in patients with estrogen receptor positive, node negative breast cancer, and early-stage node-negative patients account for about 60% of all primary breast cancer patients. The company also announced development of a new genomic test for colon cancer at the ASCO conference that will be available in 1Q10 and will be used to determine treatment options after primary tumor excision. Genomic Health is developing additional tests for therapy guidance in patients with prostate cancer, non-small cell lung cancer, renal cancer, and melanoma.

Other leading providers of genomic/proteomic tests for guidance of breast cancer therapy include Agendia (Amsterdam, the Netherlands) and American Diagnostica, (Stamford, Connecticut), a unit of Sekisui Medical (Tokyo).

As shown in Table 3, genomic tests typically identify somewhat less than half of breast cancer patients as having a low risk of recurrence, compared to the 70% of early-stage, node-negative patients typically quoted by oncologists who do not benefit from adjuvant chemotherapy but, in absence of prognostic genomic testing, usually receive it.

Three large-scale pivotal clinical trials are now under way, as shown in Table 3, to validate the performance of genomic tests in breast cancer management. Ten-year results from the Chemo-NO trial of the American Diagnostica Femtelle uPA/PAI-1 assay in breast cancer patients, discussed at the ASCO meeting by Nadia Harbeck, MD, of Frauenklinik der Technischen Universit t (Munich, Germany), show an 87% disease-free survival rate for patients who were classified as low risk by the uPA/PAI-1 test and did not receive adjuvant chemotherapy. A more recent follow-on study, which employed taxanes and Herceptin therapy when indicated, found that 39% of patients can be spared from undergoing adjuvant chemotherapy with no significant decrease in disease-free survival or overall survival.

The tests are not perfect, however, as indicated by data from a meta-analysis of studies of the Agendia MammoPrint assay, discussed by Richard Bender of Agendia at the conference, which showed that patients classified as low risk who received adjuvant chemotherapy have a disease-free survival rate at five years of 98% vs. 93% for low-risk patients who do not undergo chemotherapy. That difference was not statistically significant, but nevertheless needs to be addressed, according to Claudine Isaacs, MD, of Lombardi Comprehensive Cancer Center (Washington), who discussed the MammoPrint trial results at ASCO.

Genomic profiling tests for lung cancer could potentially address an even larger patient population than for breast cancer based on global cancer incidence figures. A genomic profiling test for non-small cell lung cancer is in fact one of the tests in the development pipeline for Genomic Health, the current leader in cancer genomic testing.

Gene expression profiling in lung cancer has, however, proven to be challenging, as discussed by Neil Hayes, MD, of the University of North Carolina (Chapel Hill) at the conference. There is a significant need for improved technologies for therapy guidance in lung cancer, since some early stage (Stage I) patients who are generally not treated with chemotherapy have been shown to benefit significantly from adjuvant chemotherapy, while other patients with more advanced disease may not benefit and could be spared from toxic treatment.

Hayes described a clinical trial now under way, designated CALGB 30506, employing microarray genomic profiling (the Lung Metagene Score) to evaluate patients and compare outcomes for patients treated with different chemotherapy regimens versus their tumor's genomic profile. Initially, a 44,000-gene profile was developed, but it is now being condensed to a smaller signature, perhaps to as few as five genes.

Certain protein markers that have been studied for a number of years in lung cancer, including TTF1, vimentin, and SMA, appear to give results that are equivalent to those obtained with genomic profiling, but the markers must be measured by immunohistochemical methods which are often not highly reproducible.

Hayes described development of a nucleic acid-based gene expression profile which is expected to allow more precise data to be obtained, eventually providing an improved capability to select chemotherapeutic as well as biologically targeted agents for lung cancer.

Another group of tests becoming increasingly important in cancer therapy is pharmacogenomic tests used to identify patients whose tumors will respond to molecularly targeted drugs. One of the first molecular pharmacogenomic tests to become widely accepted was the HercepTest, an immunohistochemical assay available as an FDA-approved test kit from Dako (Glostrup, Denmark). Dako also markets the HER2 FISH pharmDx kit, employing in situ hybridization technology.

A FISH-based assay for HER2 also is available from the Vysis unit of Abbott Molecular Diagnostics (Abbott Park, Illinois).

Pharmacogenomic tests are designed to be used in tandem with a targeted drug which attacks the same target that is detected by the test. In the case of HER2, essentially all breast cancer patients in developed countries are now tested for the presence of the marker when first diagnosed to determine if their cancer is of the aggressive HER2-positive type, which will respond to treatment with Herceptin, a drug from Genentech/Roche (South San Francisco, California) which targets the HER2 receptor.

At the ASCO meeting, a number of promising new applications for Herceptin were described which could expand the use of the drug and drive additional demand for HER2 testing. In the Phase III ToGA trial, discussed at the conference by Eric Van Cutsem, MD, PhD, of University Hospital Gasthuisberg (Leuven, Belgium), the addition of Herceptin to standard chemotherapy for patients with HER2-positive gastric cancer resulted in a significant increase in median survival, from 11.1 to 13.8 months.

Gastric (stomach) cancer has the fourth-highest incidence worldwide, with more than 1 million new cases per year.HER2 positivity is similar to that for breast cancer, i.e., about 30% of patients. The trial screened 3,800 patients, and enrolled 584 who were positive for the HER2 marker by either immunohistochemistry or FISH testing.

The patients were randomized to receive either standard chemotherapy (290) or chemotherapy plus Herceptin (294). A high response rate of 34.5% to 47.3% was achieved, and there was no difference in side effects between the two arms. In particular, there was no increase in cardiac toxicity in the Herceptin-treated group, as has been observed in breast cancer patients. A 26% reduction in risk of death was reported, marking the first time a biological therapy has been shown to provide a survival benefit in gastric cancer.

Another advance for Herceptin therapy was reported by Jose Baselga, MD, of Vall d'Hebron University Hospital (Barcelona, Spain). The T-DMI study was a 112-patient Phase II trial that evaluated the new antibody-drug conjugate T-DMI (trastuzumab-DMI) in HER2-positive breast cancer. T-DMI combines Herceptin with the chemotherapy agent DM1, using the targeting capability of Herceptin to not only inactivate the HER2 receptor but also to concentrate DM1 at the site of the tumor cell. A 44% response rate was achieved, and a Phase III trial, the EMILIA study, is now in progress in patients whose breast cancer has progressed following previous anti-HER2 therapy.

Advanced tests for HER2 were described at the ASCO conference by Monogram Biosciences (South San Francisco, California). The HERmark Breast Cancer Assay uses two monoclonal antibodies specific for unique epitopes on the HER2 receptor, and employs Monogram's VeraTag fluorescent reporter technology, which provides a significant increase in sensitivity and specificity compared to conventional immunohistochemical HER2 assays.

The test is performed as a reference lab assay by Monogram Biosciences on paraffin-embedded specimens with a 5-7 day turnaround time. Cost is $3,400 per test. A study using the HERmark test to re-analyze breast tumor specimens which had been initially assayed using FISH found that 25% of the specimens were reclassified. In addition, the test's improved performance may enable more tumor types to be identified which express HER2, further expanding the use of Herceptin and of HER2 testing.

Monogram also is developing a test for HER3 total protein, the H3T test, which has potential applications in monitoring of response to Herceptin therapy. The H3T test also uses the VeraTag technology. Studies have demonstrated that levels of HER3 expression in breast tumors as measured by the H3T test correlate with time to progression, indicating the test could be useful for identification of a population of patients who require additional therapy. The assay is now entering clinical evaluation.

A number of additional molecular targets have now been identified for anti-cancer drugs, including IGF-1R (insulin-like growth factor 1 receptor), a transmembrane protein that is over-expressed in many tumor types including lung, prostate, pancreas and breast cancer, and which mediates proliferation, metastasis, and apoptosis.

More than 20 drugs that target IGF-1R are in development by companies including Hoffman-La Roche (Basel, Switzerland), Amgen (Thousand Oaks, California), Novartis (Basel, Switzerland), Bristol-Myers Squibb (Princeton, New Jersey), Merck Oncology (North Wales, Pennsylvania), Biogen Idec (Cambridge, Massachusetts), OSI Pharmaceuticals (New York), and AntiSense Therapeutics (Victoria, Australia).

Tests for IGF-1R have been developed by the Ventana Medical Systems (Tucson, Arizona) unit of Roche and Pathogenesys (San Juan Capistrano, California). The Ventana test employs a new technology, silver in situ hybridization, an alternative to FISH that offers the advantage of permanent archiving in contrast to FISH specimens which deteriorate over time.

An evaluation of the Ventana test using archival lung tumor specimens to predict outcome found that the test had a better predictive power than histology, and also outperformed conventional staging except for early versus late-stage disease.

Other important drug targets in oncology for which drugs and companion tests are under development include K-RAS, ERCC1, B-RAF, RRM1, TS and EGFR. One company, Response Genetics (Los Angeles), has developed a series of test panels for a number of key drug targets which can be used to guide therapy for lung and colon cancer. The company's ResponseDX:Colon panel, for example, measures expression of EGFR (epidermal growth factor receptor) in biopsy specimens as an indication of sensitivity to Tarceva (erlotinib), a targeted drug from OSI Pharmaceuticals and Genentech/Roche used in the treatment of lung and colorectal cancer.

Other drugs which target EGFR include Erbitux (cetuximab) from Bristol-Myers Squibb and Vectibix (panitumumab) from Amgen. K-RAS is another marker which is useful in predicting response to EGFR-targeted drugs. The panel also detects K-RAS, excision repair cross-complementing factor 1 (ERCC1), ribonucleotide reductase M1 (RRM1), thymidylate synthase (TS), and vascular endothelial growth factor receptor 2 (VEGFR2) expression. ERCC1, RRM1, and TS are indicators of sensitivity to platin-based chemotherapy, gemcitabine, and FOLFOX chemotherapy, while VEGFR2 expression is associated with sensitivity to anti-angiogenic therapies such as Genentech/Roche's Avastin (bevacizumab).

Genzyme Genetics (Cambridge, Massachusetts) is another provider of testing services that offers personalized tests for drug targets such as ERCC1, KRAS, and EGFR. Applied Genomics (Huntsville, Alabama) sells an immunohistochemistry test kit, MammoTAX, a single antibody test for determining whether a patient will respond to taxane therapy based on presence of the TLE3 marker.

Taxane drugs include Taxol (paclitaxel) from Bristol-Myers Squibb and Taxotere (docetaxel) from Sanofi-Aventis (Paris).

Another emerging company in the cancer drug marker testing segment is Expression Pathology (Rockville, Maryland), which has developed and patented Liquid Tissue proteomics technology for analyzing protein biomarkers of drug response and disease progression. The company recently received $6.5 million in funding from Abraxis Bioscience (Los Angeles), a leader in the targeted anticancer drug field, to develop proteomic clinical assays.

One of the most notable advances in targeted cancer therapy announced at the ASCO meeting was results from a trial of a new class of drugs called PARP inhibitors in the treatment of triple negative breast cancer. PARP inhibitors target the DNA repair enzyme poly ADP-ribose polymerase, reducing the tumor cell's ability to resist chemotherapy agents, and thereby increasing the effectiveness of therapy.

PARP has been found to be elevated in certain difficult-to-treat cancers such as BRCA 1/2-positive breast cancer, including triple-negative breast cancer. Furthermore, in BRCA 1/2 tumors, the cells are believed to be solely dependent on PARP for DNA repair, whereas normal cells have additional mechanisms available such as homologous recombination. Consequently, a PARP inhibitor should only target tumor cells, and leave normal tissues unaffected.

Results from a trial of a new PARP inhibitor, BSI-201, a drug developed by BiPar Sciences, recently acquired by Sanofi-Aventis, was described at an ASCO press conference by Joyce O'Shaughnessy, MD, of Baylor Sammons Cancer Center and US Oncology (both Dallas). The trial enrolled patients with triple-negative breast cancer, which is negative for estrogen and progesterone receptors as well as HER2.

Triple negative breast cancer comprises 15% of all breast cancer, or about 170,000 cases annually worldwide according to O'Shaughnessy. Almost all tumors which carry the BRCA1 mutation are triple-negative. Results from the trial of BSI-201 showed a doubling in progression-free survival, from 3.7 to 6.9 months, and an increase in overall survival from 5.7 to 9.2 months, corresponding to a 60% reduction in death rate.

BSI-201 is the first targeted agent for treatment of triple-negative breast cancer, and will next be studied in a Phase III trial that started in June.

AstraZeneca also is developing a PARP inhibitor, AZD2281, an oral agent which is being evaluated in a Phase II trial. Results from a pilot study of the drug were discussed at ASCO by Andrew Tutt, MD, of Guys Hospital (London). A 41% overall response rate was reported, with one complete response and 10 partial responses. The response rate was dose-dependent, and the drug was well-tolerated.

Myriad Genetics (Salt Lake City) is the exclusive provider of BRCA testing services, and is performing BRCA testing for patients enrolled in the AZD2281 trial.

Therapy Optimization Is Key Trend

In addition to using genomic and molecular testing to select the most effective drug therapy for individual cancer patients, personalized cancer treatment also involves optimization of the therapeutic strategy based on each patient's disease characteristics and status. Therapy optimization can include improved approaches to delivering an anti-cancer drug to a tumor, as well as altering drug dose or delivery schedule based on a balance of minimizing toxicity versus maximizing therapeutic effect.

Pharmacogenomics also plays a key role in therapy optimization by enabling an individual patient's response to drug therapy to be predicted prior to delivery of the drug, and the related area of therapeutic drug monitoring in oncology enables optimal drug dosing based on each patient's ability to metabolize, absorb, and excrete the drug. As an example of the latter technology, Myriad Genetics has recently introduced a new product, OnDose, which is designed for use in optimizing the dose of 5-fluorouracil used in chemotherapy.

OnDose was launched in April, and is a test performed at Myriad's laboratory in Salt Lake City at a cost of $300 per test. The test is performed once per 5-FU dose cycle, and measures the level of the drug in the patient's blood. A study comparing 5-FU treatment guided by OnDose testing vs. dose selection based on body surface area (BSA) showed a 33.7% response rate for the OnDose cohort vs. 17.3% for the BSA group. Toxicity occurred in 4% of OnDose patients vs. 18% in the BSA cohort. Myriad provides a specimen collection and transportation kit for the test.

The company also offers the TheraGuide 5-FU test, a pharmacogenomic test which measures variations in the DPYD and TYMS genes that are responsible for a significant portion of serious adverse reactions to 5-FU therapy. The TheraGuide test can be used in conjunction with OnDose to determine both the optimal initial dose for 5-FU therapy and follow-on dose levels during treatment.

Myriad estimates that about 175,000 patients in the U.S. are candidates for OnDose testing to optimize 5-FU therapy in colorectal cancer. Even though OnDose was only introduced in April, the test already is reimbursed by insurers covering over 100 million lives in the U.S.

A new approach for optimizing treatment of ovarian cancer was described at ASCO by Gordon Rustin, MD, of Mount Vernon Cancer Centre (London). The existing standard of care for management of ovarian cancer following primary treatment is to monitor levels of the CA125 tumor marker at three-month intervals to check for recurrence, and to initiate chemotherapy when a rise in CA125 is detected.

Rustin evaluated an alternative strategy using onset of clinical symptoms to detect relapse, with CA-125 testing performed only as needed. The study involved 1,442 patients in 10 countries, and spanned a nine-year interval beginning in 1996. Some 37% of the patients initially screened were randomized to either early therapy initiated based solely on a rise in CA125 (265) or delayed therapy based on clinical symptoms of recurrence (264).

The trial was unusual in that a very high five-year survival rate of 70% was observed in the overall patient population. The time to beginning of chemotherapy in the CA125-monitored group was 0.8 months following detection of relapse, vs. 5.6 months in the symptom-guided group. There was no difference in overall survival for the two groups, and global health scores were worse for the CA125 group. Quality of life also was not improved with CA125 testing.

Rustin said the results show that routine CA125 testing to detect recurrence of ovarian cancer is not beneficial in extending survival or improving quality of life, and furthermore that testing causes great anxiety for patients. The results potentially will alter the standard of care for ovarian cancer patients, reducing costs related to routine CA125 tests as well as other diagnostic procedures such as CT or MR scans performed to verify recurrence, while improving quality of life.

CA125 testing is typically performed in the hospital central lab or a reference laboratory using test kits supplied by Abbott Diagnostics, Ortho Clinical Diagnostics (Raritan, New Jersey), Beckman Coulter (Brea, California), and Roche Diagnostics (Indianapolis), among others.

In a follow-on discussion of the results presented by Rustin, Beth Karlan, MD, of Cedars Sinai Medical Center (Los Angeles), stated that improved methods for early detection of primary ovarian cancer are needed to enable more effective treatment, minimizing the rate of recurrence, and that many targeted therapies are now in development which may prove to be more effective in treating early-stage cancer such as that detected with CA125.

Another study involving optimization of cancer therapy was presented at ASCO by Robert Epstein, MD, of Indiana University School of Medicine (Indianapolis). The study evaluated the effect of concomitant therapy with drugs which inhibit the CYP2D6 enzyme and tamoxifen on breast cancer recurrence.

Tamoxifen must be converted to an active form in the body in order to be effective, and conversion is dependent on activity of CYP2D6. A number of commonly prescribed drugs such as the anti-depressants Zoloft and Prozac inhibit CYP2D6, and would be expected to cause a drop in effectiveness of Tamoxifen.

Epstein reported a 13.9% recurrence rate for breast cancer in patients who were taking a moderate-to-potent CYP2D6 inhibitor vs. a 7.5% rate for those not taking any such drugs. The study employed the 10.7 million-member Medco database to obtain data on drug use. CYP2D6 inhibitors are commonly prescribed in up to 30% of breast cancer patients in the U.S., primarily in pre-menopausal women.

Conversely, a second study which addressed the same topic performed at Leiden University Medical Center in the Netherlands found no effect of CYP2D6 inhibitor use on the effectiveness of tamoxifen in preventing breast cancer recurrence. As reported by Vincent Dezentje, MD, at the conference, adherence to tamoxifen therapy was poor in the Leiden trial, but CYP2D6 inhibitor use did not adversely affect the activity of tamoxifen.

Dezentje noted that aromatase inhibitors have largely replaced tamoxifen for post-menopausal breast cancer patients in the U.S. and Europe, but that the drug continues to be used in other regions as well as for pre-menopausal patients. He recommended that, in light of the results from the U.S. study described by Epstein, other types of anti-depressants be prescribed which do not inhibit enzymes such as CYP2D6.

Early detection, prevention remain goal

While optimization of therapy as well as personalized cancer treatments are clearly desirable and beneficial advances in cancer treatment, the most effective strategy for reducing deaths from cancer is early detection and, ultimately, prevention of the disease. As discussed by Bert Vogelstein, MD, of Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins University (Baltimore), in a plenary lecture at ASCO, the timeframe over which the 50 to 100 genetic mutations occur which are typically required for cancer development is approximately 30 years.

Today, any patient who dies from colorectal cancer does so only because the disease was not detected in the 27 years when it was developing. If detected early, colorectal cancer as well as many other major types of cancer can be cured with surgery, radiotherapy and chemotherapy.

While early detection is not the norm for most cancers today, cancer death rates have been reduced in some cases in which effective screening strategies have been widely implemented, such as for cervical cancer and breast cancer in the U.S. and other developed countries.

Vogelstein said that technologies for improving the early detection of cancer are now becoming available, including new types of molecular imaging techniques which surpass the capabilities of FDG-PET imaging, as well as advanced methods for detection of genetic mutations leading to cancer.

Vogelstein is developing a new technology, digital PCR, which involves counting individual DNA molecules, and shows promise for detection of colorectal cancer at the pre-malignant stage. Additional new early detection technologies now under development are described in Table 4.

Existing cancer screening technologies, such as colonoscopy exams combined with fecal occult blood testing, have had a positive impact on cancer death rates, lowering deaths from colorectal cancer from 57,158 in 1990 to 49,960 in 2008. Technologies for improving cancer screening, such as molecular tests for detecting genetic mutations in stool specimens related to colorectal cancer, have been introduced such as the ColoSure test developed by Exact Sciences (Marlborough, Massachusetts) and provided as a reference lab test by LabCorp (Burlington, North Carolina).

However, uptake has been limited due to current recommendations that do not include a role for fecal DNA testing, even though sensitivity of the test is 51.6% compared to only 12.9% for fecal occult blood. Molecular screening tests may be improved by implementing advanced technologies such as digital PCR or epigenetic analysis, as exemplified by the Septin 9 test now under development by Epigenomics.

Ultimately, routine screening of individuals determined to be at risk for cancer development based on family history or, perhaps, genomic analysis to track genetic mutations as they occur could enable most adult cancers to be detected while still highly curable. Given that one of every two men and one of three women in the U.S. have a lifetime risk of developing cancer, the target population for effective cancer screening tests numbers in the hundreds of millions, creating a significant market opportunity for companies in the sector.