The NEW ENGLA ND JOURNAL of MEDICINE

Perspective july 22, 2010

The Path to Personalized Medicine

Margaret A. Hamburg, M.D., and Francis S. Collins, M.D., Ph.D.

M ajor investments in basic science have created

an opportunity for significant progress in
clinical medicine. Researchers have discovered
the NIH and the FDA will invest
in advancing translational and
regulatory science, better define
regulatory pathways for coordinat-
hundreds of genes that harbor variations contributing ed approval of codeveloped diag-
nostics and therapeutics, develop
to human illness, identified ge- focusing on the best ways to de- risk-based approaches for appro-
netic variability in patients’ re- velop new therapies and optimize priate review of diagnostics to
sponses to dozens of treatments, prescribing by steering patients more accurately assess their va-
and begun to target the molecu- to the right drug at the right lidity and clinical utility, and make
lar causes of some diseases. In dose at the right time. information about tests readily
addition, scientists are developing We recognize that myriad ob- available.
and using diagnostic tests based stacles must be overcome to Moving from concept to clini-
on genetics or other molecular achieve these goals. These include cal use requires basic, translation-
mechanisms to better predict scientific challenges, such as de- al, and regulatory science. On the
patients’ responses to targeted termining which genetic markers basic-science front, studies are
therapy. have the most clinical signifi- identifying many genetic varia-
The challenge is to deliver the cance, limiting the off-target ef- tions underlying the risks of both
benefits of this work to patients. fects of gene-based therapies, and rare and common diseases. These
As the leaders of the National conducting clinical studies to newly discovered genes, proteins,
Institutes of Health (NIH) and identify genetic variants that are and pathways can represent pow-
the Food and Drug Administra- correlated with a drug response. erful new drug targets, but cur-
tion (FDA), we have a shared vi- There are also policy challenges, rently there is insufficient evi-
sion of personalized medicine such as finding a level of regu- dence of a downstream market
and the scientific and regulatory lation for genetic tests that both to entice the private sector to ex-
structure needed to support its protects patients and encourages plore most of them. To fill that
growth. Together, we have been innovation. To make progress, void, the NIH and the FDA will

n engl j med 363;4 nejm.org july 22, 2010 301

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org on November 20, 2024. For personal use only.
No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.
PERS PE C T IV E The Path to Personalized Medicine

develop a more integrated path- Heart Study, obtaining biologic opportunity for early, informal
way that connects all the steps specimens from babies enrolled feedback that may assist compa-
between the identification of a in the National Children’s Study, nies in reaching important stra-
potential therapeutic target by and performing detailed genetic tegic decisions. The goal is to
academic researchers and the analysis of 20 types of tumors help companies integrate geno-
approval of a therapy for clinical to improve our understanding of mics into their clinical-develop-
use. This pathway will include their molecular basis. ment plans.
NIH-supported centers where re- As for translational science, Today, about 10% of labels
searchers can screen thousands of the NIH is harnessing the talents for FDA-approved drugs contain
chemicals to find potential drug and strengths of its Clinical and pharmacogenomic information —
candidates, as well as public– Translational Sciences Award pro- a substantial increase since the
private partnerships to help move gram, which currently funds 46 1990s but hardly the limit of the
candidate compounds into com- centers and has awardees in 26 possibilities for this aspect of
mercial development. states, and its Mark O. Hatfield personalized medicine.1 There
The NIH will implement this Clinical Research Center (the coun- has been an explosion in the
strategy through such efforts as try’s largest research hospital, in number of validated markers but
the Therapeutics for Rare and Bethesda, MD) to translate basic relatively little independent analy-
Neglected Diseases (TRND) pro- research findings into clinical ap- sis of the validity of the tests
gram. With an open environment, plications. Just as the NIH served used to identify them in biologic
permitting the involvement of all as an initial home for human specimens.
the world’s top experts on a given gene therapy, the Hatfield Center The success of personalized
disease, the TRND program will can provide specialized diagnos- medicine depends on having ac-
enable certain promising com- tic services for rare and neglected curate diagnostic tests that iden-
pounds to be taken through the diseases, offer a state-of-the-art tify patients who can benefit
preclinical development phase — manufacturing facility for novel from targeted therapies. For ex-
a time-consuming, high-risk phase therapies, and pioneer clinical ample, clinicians now commonly
that pharmaceutical firms call trials of other innovative biologic use diagnostics to determine
“the valley of death.” Besides ac- therapies, such as those using which breast tumors overexpress
celerating the development of human embryonic stem cells or the human epidermal growth fac-
drugs to treat rare and neglected induced pluripotent stem cells. tor receptor type 2 (HER2), which
diseases, the TRND program may As genetics researchers gener- is associated with a worse prog-
also help to identify molecularly ate enormous amounts of new nosis but also predicts a better re-
distinct subtypes of some com- information, the FDA is develop- sponse to the medication tras­
mon diseases, which may lead to ing the regulatory science stan- tuzumab. A test for HER2 was
new therapeutic possibilities, ei- dards and evidence needed to approved along with the drug (as
ther through the development of use genetic information in drug a “companion diagnostic”) so that
targeted drugs or the salvaging and device development and clin- clinicians can better target pa-
of abandoned or failed drugs by ical decision making. The agen- tients’ treatment (see table).
identifying subgroups of patients cy’s Critical Path Initiative aims Increasingly, however, the use
likely to benefit from them. to develop better evaluation tools, of therapeutic innovations for a
Another important step will be such as biomarkers and new as- specific patient is contingent on
expanding efforts to develop tis- says. Under the Voluntary Geno­ or guided by the results from a
sue banks containing specimens mic Data Submission program, diagnostic test that has not been
along with information linking companies can discuss genetic independently reviewed for accu-
them to clinical outcomes. Such information with the FDA in a racy and reliability by the FDA.
a resource will allow for a much forum separate from the product- For example, in 2006, the FDA
broader assessment of the clini- review process. These discussions granted approval to rituximab
cal importance of genetic varia- give the agency and companies a (Rituxan) for use as part of first-
tion across a range of conditions. better understanding of the sci- line treatment in patients with
For example, the NIH is now sup- entific issues involved in applying certain cancers. Since then, a lab-
porting genome analysis in par- pharmacogenomic information to oratory has marketed a test with
ticipants in the Framingham drug development and offer an the claim that it can identify the

302 n engl j med 363;4 nejm.org july 22, 2010

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org on November 20, 2024. For personal use only.
No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.
PE R S PE C T IV E The Path to Personalized Medicine

Examples of FDA-Approved Drugs and Companion Diagnostics in Clinical Practice.*

Approved Drug Mechanism Approved Companion Diagnostic

Herceptin (trastuzumab) Targets HER2 to treat metastatic breast HER2 immunohistochemistry tests, HER2
cancer gene-amplification tests
Erbitux (cetuximab) Targets EGFR to treat metastatic colorectal EGFR immunohistochemistry test
cancer
Gleevec (imatinib) Targets the cell-surface tyrosine kinase re- c-kit immunohistochemistry test
ceptor c-kit in gastrointestinal stromal
tumors

* EGFR denotes epidermal growth factor receptor, and HER2 human epidermal growth factor receptor type 2.

approximately 20% of patients The agency’s goal is an efficient cancer; an ovarian cancer test,
who are more likely to have a review process that produces di- marketed before the completion
response to the drug. The FDA agnostic–therapeutic approaches of an NIH-funded study,2 gave
has not reviewed the scientific that clinicians can rely on and false readings that reportedly led
justification for this claim, but allows companies that invest in to the unnecessary removal of
health care providers may use establishing the validity and use- women’s ovaries; and flawed,
the test results to guide therapy. fulness of tests to make specific, mishandled data underlying a
This undermines the approval FDA-backed claims about benefits. test for Down’s syndrome were
process that has been established Patients should be confident discovered only days before the
to protect patients, fails to en- that diagnostic tests reliably give test was to go on the market.
sure that physicians have accu- correct results — especially when Through a process that includes
rate information on which to test results are used in making opportunities for public input,
make treatment decisions, and major medical decisions. The FDA the FDA will work to ensure the
decreases the chances that phy- has long taken a risk-based ap- quality of key diagnostic tests,
sicians will adopt a new thera- proach to the oversight of diag- helping to protect patients and
peutic–diagnostic approach. The nostic tests, historically focus- giving clinicians confidence that
FDA is coordinating and clarify- ing on test kits that are broadly personalized medicine will lead
ing the process that manufactur- marketed to laboratories or the to real health improvements.
ers must follow regarding their public (e.g., pregnancy tests or In addition, the NIH will ad-
claims, including defining the blood glucose tests); such kits are dress the fact that there is no
times when a companion diagnos- sold only if the FDA has deter- single public source of compre-
tic must be approved or cleared mined that they accurately provide hensive information about the
before or concurrently with ap- clinically significant information. more than 2000 genetic tests
proval of the therapy. The agency But recently, many laboratories that are available through clini-
will ensure that claims that a have begun performing and broad- cal laboratories. On the recom-
test will improve the care of pa- ly marketing laboratory-developed mendation of a federal advisory
tients are based on solid evi- tests, including complicated ge- committee,3,4 the NIH — with
dence, and developers will get netic tests. The results of these advice from the FDA, other De-
straightforward, consistent advice tests can be quite challenging to partment of Health and Human
about the standards for review interpret. Because clinicians may Services agencies, and diverse
and the best way to demonstrate order a genetic test only once, stakeholders — is creating a vol-
that the combination works as getting the results right the first untary genetic testing registry
intended. time is crucial. to address key information gaps.5
Genetic tests are not perfect, There are reports of problems Readily available information about
in part because most gene muta- with laboratory tests that have these tests, including whether
tions do not perfectly predict out- not had FDA oversight: women they were cleared or approved by
comes. Clinicians will need to were erroneously told they were the FDA, will help clinicians and
understand the specificity and negative for a mutation confer- consumers make informed deci-
sensitivity of new diagnostics. ring a very high risk of breast sions about using the tests to op-

n engl j med 363;4 nejm.org july 22, 2010 303

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org on November 20, 2024. For personal use only.
No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.
PERS PE C T IV E The Path to Personalized Medicine

timize health care. The registry standing of the genetic basis of This article (10.1056/NEJMp1006304) was
published on June 15, 2010, and updated on
will also support scientific dis- disease. We also anticipate that September 8, 2010, at NEJM.org.
coveries by facilitating the shar- some previously failed medica-
ing of data about genetic variants. tions will be recognized as safe 1. Frueh FW, Amur S, Mummaneni P, et al.
In February, the NIH and the and effective and will be ap- Pharmacogenomic biomarker information
in drug labels approved by the United States
FDA announced a new collabora- proved for subgroups of patients Food and Drug Administration: prevalence
tion on regulatory and transla- with specific genetic markers. of related drug use. Pharmacotherapy 2008;
tional science to accelerate the When the federal government 28:992-8.
2. Ovarian cancer research results from the
translation of research into med- created the national highway sys- Prostate, Lung, Colorectal and Ovarian
ical products and therapies; this tem, it did not tell people where (PLCO) Cancer Screening Trial: fact sheet.
effort includes a joint funding to drive — it built the roads and Bethesda, MD: National Cancer Institute.
(Accessed June 11, 2010, at http://www.cancer
opportunity for regulatory science. set the standards for safety. Those .gov/cancertopics/cancertopics/factsheet/
Working with academic experts, investments supported a revolu- detection/plco-ovarian.)
companies, doctors, patients, and tion in transportation, commerce, 3. Secretary’s Advisory Committee on Ge-
netic Testing. Enhancing the oversight of ge-
the public, we intend to help and personal mobility. We are now netic tests: recommendations of the SACGT.
make personalized medicine a building a national highway sys- Bethesda, MD: National Institutes of Health,
reality. A recent example of this tem for personalized medicine, 2000. (Accessed June 11, 2010, at http://oba
.od.nih.gov/oba/sacgt/reports/oversight_
collaboration is an effort to iden- with substantial investments in report.pdf.)
tify new investigational agents infrastructure and standards. We 4. Secretary’s Advisory Committee on Ge-
to which certain tumors, identi- look forward to doctors’ and pa- netics, Health, and Society. U.S. system of
oversight of genetic testing: a response to the
fied by their genetic signatures, tients’ navigating these roads to charge of the Secretary of Health and Human
are responsive. better outcomes and better health. Services. Bethesda, MD: National Institutes
Real progress will come when Disclosure forms provided by the au- of Health, 2008. (Accessed June 11, 2010, at
thors are available with the full text of this http://oba.od.nih.gov/oba/SACGHS/reports/
clinically beneficial new prod- SACGHS_oversight_report.pdf.)
article at NEJM.org.
ucts and approaches are incor- 5. Genetic Testing Registry. Bethesda, MD:
porated into clinical practice. As Dr. Hamburg is the commissioner of the National Center for Biotechnology, National
Food and Drug Administration, Silver Library of Medicine; 2010. (Accessed June 11,
the field advances, we expect to 2010, at http://www.ncbi.nlm.nih.gov/gtr.)
Spring, and Dr. Collins is the director of the
see more efficient clinical trials National Institutes of Health, Bethesda — Copyright © 2010 Massachusetts Medical Society.
based on a more thorough under- both in Maryland.

The ACA’s New Weapons against Health Care Fraud

John K. Iglehart

M arshaling expanded finan-

cial resources, aggressive
new legal authority, and rare bi-
rigorous crackdown on illegal ac-
tivities that plague Medicare,
Medicaid, and private insurers.
vulnerable to fraud, waste, and
abuse. In 2009, government-wide
“improper payments” totaled $98
partisan solidarity, the Obama Under past policies, Congress and billion — more than half of it
administration is accelerating fed- the executive branch “way, way, paid by Medicare and Medicaid.
eral efforts to fight health care way” underspent on fighting It is uncertain how much of the
fraud, waste, and abuse that cost health care fraud, according to activity that resulted in these pay-
taxpayers and private insurers bil- Kerry Weems, who was acting ments was actual fraud, but Lewis
lions of dollars every year. Al- administrator of the Centers for Morris, chief counsel of the Of-
though the new forms of author- Medicare and Medicaid Services fice of Inspector General (OIG),
ity are granted by the Affordable (CMS) from 2007 to 2009.2 Department of Health and Hu-
Care Act (ACA), which Republi- Since 1990, the Government man Services (DHHS), told the
cans unanimously opposed, most Accountability Office (GAO) has Senate Finance Committee last
GOP legislators strongly support designated Medicare as a high- year, “Although we cannot mea-
— and some even sponsored risk federal program because its sure the full extent of health
measures to enable1 — the more vast size and complexity make it care fraud in Medicare and Med-

304 n engl j med 363;4 nejm.org july 22, 2010

The New England Journal of Medicine is produced by NEJM Group, a division of the Massachusetts Medical Society.
Downloaded from nejm.org on November 20, 2024. For personal use only.
No other uses without permission. Copyright © 2010 Massachusetts Medical Society. All rights reserved.