Personalized Medicine Complicates Medical Decision-Making

As the use of biomarkers grows, managed care companies will face tough decisions about setting limits

Four things that I read recently really struck me as the deadline for “Tomorrow’s Medicine” approached.

First, the new recommendations made by the American Society of Clinical Oncology (ASCO) in the 2007 Update of Recommendations for the Use of Tumor Markers in Breast Cancer, which reviewed the available technology in tumor markers and its use in prevention screening, treatment, and surveillance of breast cancer. Thirteen categories were considered, six of which were new since 2000.

Second, a news article in which a major presidential candidate vilified a leading health plan for failing to approve an organ transplant for a patient who later died.

Third, the United States Food and Drug Administration’s Draft Guidance for Industry, Clinical Laboratories, and FDA Staff: In Vitro Diagnostic Multivariate Index Assays, which discussed the combination of values of multiple variables using a function to yield a single, patient-specific result or score that is intended to be used in the diagnosis or treatment decisions for a specific disease.

Finally, an article in the Wall Street Journal titled “Bad Cancer Tests Drawing Scrutiny,” which discussed the variability of reported results when a single tissue sample is analyzed by two separate labs and the results conflict.


These four publications highlight the difficult challenges facing health care as it increasingly incorporates a personalized approach that uses various biomarkers to influence medical decision-making.

In the purest sense, a biomarker is a biologic measurement. Biomarkers may be images, physical measurements, tissue samples, or chemical or biologic tests performed using blood or serum. They are used to measure risk — for screening, diagnosis, prognosis, and sensitivity to medications.

Recent integration of computer technology with complex biochemistry and a better understanding of cellular processes have led to an explosion of new tests using biomarkers in microarrays. The number of biomarkers available to physicians and patients is increasing quickly as more cell signaling pathways are discovered and, eventually, used in commercial testing.

In addition, the use of combined biomarkers to determine a composite score is also increasing as evidenced by the recent recommendation by ASCO to use Oncotype DX, a product created by Genomic Health, to aid in adjuvant care decision making for patients with breast cancer.

These advances are occurring in an environment where the FDA has effectively limited its enforcement authority, where patients are demanding more, and where employers and health plans are struggling to cover even basic care, all the while attempting to determine a cost-effective approach to diseases whose complexity is increasingly apparent.

As our knowledge about cell signaling — those complex communication pathways between and within cells — advances, we are rewriting medical textbooks almost daily.

Many of these pathways are known to be pathologically activated or deactivated in the presence of disease, particularly cancerous growths. It is hoped that some component of these pathways will eventually become a biomarker for the disease.

This explosion of knowledge about basic cell metabolism has made decision-making even more complex when a patient is faced with a life-threatening or life-limiting disease. In addition, our governmental regulatory agencies, medical training programs, and payment processes are poorly equipped to deal with this complexity.

Limited resources

Organizations such as ASCO must be applauded for their efforts to develop recommendations based upon the best science available.

Health plans have looked to professional societies in determining policy, but the last time ASCO updated its guidelines was seven years ago. In this digital age, how can a finite organization possibly maintain up-to-date policies to assist in the making of complex medical treatment and payment decisions? In addition, biomarkers require validation. The FDA is attempting to provide guidance in this arena, but it does not have sufficient resources.

The use of biomarkers has other limitations. Recently, the BlueCross BlueShield Technology Evaluation Center released an assessment titled “Gene Expression Profiling of Breast Cancer to Select Women for Adjuvant Chemotherapy.”

The authors cited several limitations: “Among those willing to be guided by the test result, it is unknown what proportion of conventionally estimated intermediate- to-high risk patients will have sufficiently low risk score values to change their decision regarding chemotherapy.” Further, “the recurrence risk level below which women are comfortable without chemotherapy is unknown.”

And “because the risk score is a continuous function with respect to recurrence rates, risk category cutoff values selected by the test developers are arbitrary and may not be optimal.”

Managed care companies will increasingly need highly trained professionals to sort through all of these issues as Tomorrow’s Medicine continues to discover the wonder of life at the cellular and molecular level.

More questions than answers

Scientific discoveries often lead to some answers, but new questions arise as well.

  • Should the desired positive and negative predictive value differ when safety and effectiveness are considered separately?
  • Will every test be standardized?
  • Should only high-volume labs be allowed to perform tests?
  • Should labs be certified? If so, by whom?
  • Should only certified tests or labs be paid?
  • What will government payers do?
  • What incentives were involved in the development or use of the tests and can the incentives be quantified and will they be transparent?
  • Should drug manufacturers be required to provide a validated biomarker for their drugs?
  • What happens when a test predicts an efficacy for a given medication that is 1 percentage point less than what a health plan has set in policy?
  • What about subjectivity in the outcome of some tests where a cytologist or pathologist must make a visual judgment?
  • Will patients abide by the arbitrary nature of cutoff points?
  • What appeal processes will be in place to ensure due process of decisions by health plans?
  • If a payer pays for these expensive tests, should the patient be held to the result?
  • If a patient personally pays for a test that determines that a specific denied drug is indicated, how should the health plan respond?
  • If a new drug is developed that addresses the same specific receptor or biomarker as an older, more studied drug, will the new drug be seen as equal to the older one?
  • Can payers, providers, laboratories, and pharmaceutical manufacturers agree on how to use these new biomarkers?
  • Finally, how will politics treat these complex issues?
MANAGED CARE February 2008. ©MediMedia USA

Our most popular topics on