| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
GUEST EDITORIAL |
Pharmacogenetics Group, Ingenix Epidemiology, and Department of Oral Health Policy & Epidemiology, Harvard School of Dental Medicine, 275 Grove Street, Suite 3-120, Newton, MA 02466, USA; thanos.zavras{at}ingenix.com
KEY WORDS: clinical trial • genomics • safety • drug • material
The clinical practice of dental medicine uses many chemical compounds, drugs, biologics, restorative biomaterials, and devices. The richness in the armamentarium of the dental practitioner allows for therapeutic interventions that were previously unimaginable; at the same time, however, clinicians bear a great deal of responsibility for knowledge about and understanding of the relative benefits and risks of the available therapies. One area of responsibility is the voluntary reporting of any unexpected adverse effects. Such a voluntary reporting system has served our society by raising awareness around drugs, materials, or devices that increase one’s risk of developing an adverse drug reaction (ADR).
In addition to the clinical responsibility to report adverse events, the dental and oral health research community is involved in the systematic evaluation of drugs, materials, and devices. Most safety evaluation in humans occurs during pre-market clinical trials. However, clinical trials do not answer all the questions on safety and efficacy. Although they are considered the ’gold standard’ in evidence-based dentistry, one needs to understand their limitations. Clinical trials are designed to evaluate a narrow spectrum of safety and efficacy; patients are selected based on strict inclusion and exclusion criteria. Disease definition is scrutinized, and co-morbidities are usually excluded.
After the approval of a drug, material, or device, its clinical use typically extends to a broader range of afflicted individuals within the disease category, indications expand in practice, and the co-morbidities that might have excluded patients from clinical trials, as not being contraindications, may become commonplace among treated patients. The result is that drugs and biomaterials are used under conditions that have never been evaluated.
The Sept. 30, 2004, announcement by Merck to withdraw rofecoxib (Vioxx) from the US market reminded us of the many limitations in the system of post-marketing safety surveillance (Meier, 2004). In brief, the current system relies on pharmaceutical companies and the regulatory authority—the FDA in the USA, the EMEA in Europe, etc.—to monitor safety signals through registration of voluntary reports. The signals are assessed qualitatively by review and numerically against the role of chance in strata of different classes of patients. When reports accumulate and statistics point to the direction of a non-random effect, hypotheses are generated and formal epidemiologic studies are conducted. Although regulators (FDA, EMEA, etc.) are pivotal in requesting such studies, not all safety research initiates as a result of their mandate. In the United States, the nature of the FDA’s role in post-marketing safety assessment has recently been questioned, because the same office of the FDA that approves a new drug also has the overall responsibility to assess its long-term post-market safety and efficacy. Many see a potential for conflict in maintaining both functions within the same group, with a disincentive to act against a drug that the same office had previously judged to be safe (Psaty et al., 2004; Strom, 2004).
Another element of concern for the public and the scientific community is the non-disclosure or delay of disclosure of research findings that pertain to safety. Asymmetry of information, caused by non-disclosure of research by the pharmaceutical manufacturer, seems to obscure the safety profile of a drug or a biologic to the direction of the null (no significant risk). Such asymmetry has been associated with two main factors, conflict of interest in asking industry to monitor its own drugs, and lack of adequate resources to conduct independent post-marketing surveillance and pharmaco-epidemiological studies (Psaty et al., 2004; Strom, 2004). There are obvious solutions to both, but their implementation requires society’s commitment to a better, more transparent, perhaps more regulated, but certainly more expensive system of surveillance. Further, the current system relies heavily on statistical evaluation of signals (vs. chance) and the mathematical quantification of risk within the exposed; the role of biology is often neglected. In the past, such neglect was often justified. A formal study to evaluate a biologic pathway or genetic predisposition would require a large sample size of rarely occurring cases, along with the expenditure of resources to identify new or validate existing biomarkers in the laboratory, obtain specimens utilizing invasive procedures, test the clinical samples, and statistically analyze the results utilizing methods of questionable validity (due to multiple comparisons). However, several technologies exist today to overcome most of the above-named barriers. Large administrative databases of health insurance plans can be mined to identify extremely rare outcomes and enroll them in epidemiologic or pharmacogenetic studies (Walker, 2001; Zavras, 2004); microarray experiments can interpret gene expression profiles in toxicogenomic assays (Lord and Papoian, 2004); extraction of DNA is possible using non-invasive methods to collect oral epithelial cells (Harty et al., 2000); small DNA quantities can now be amplified using whole genome amplification (WGA) techniques (Hosono et al., 2003); high-throughput genotyping platforms allow for overnight testing of hundreds of thousands of polymorphisms (Cox, 2003); novel statistical methods are being developed to allow for haplotype evaluation (Goldstein et al., 2003; Hinds, 2004); and advanced data-mining techniques applied to dense genome maps can pinpoint genetic differences in the sequence or the expression of ‘high-risk’ genes. Biology must play a more significant role in post-market safety surveillance.
Genomic dental medicine, also called ‘personalized’ or ‘individualized’ dental medicine, has the potential to revolutionize our profession by predicting individual disease risk, by detecting individual response to drugs, or by designing personalized treatment plans according to one’s unique genetic characteristics. For drugs and biomaterials, the many benefits of the Human Genome Project will be mostly realized by applying its lessons to already-existing products and by ‘improving’ their use. Our challenge is to catalyze the reactions needed to raise biology to its proper place and institute it into post-marketing surveillance by means of increased appropriation of funds and Federal regulations.
ACKNOWLEDGMENTS
No conflict of interest is declared. Dr. Zavras is on a one-year leave of absence from the Medical Faculty of Harvard, currently serving as senior scientist at Ingenix Research. In this capacity, he is responsible for pharmacogenetic studies and relevant consulting to the pharmaceutical and biotech industry. The opinions expressed are those of the author.
Received December 3, 2004; Accepted December 14, 2004
REFERENCES
Cox D (2003). Finding genetic hot spots using genome-wide scans. BioIT World. Available electronically at http://www.bio-itworld.com/news/040403_report2266.html (accessed 12/01/2004).
Goldstein DB, Ahmadi KR, Weale ME, Wood NW (2003). Genome scans and candidate gene approaches in the study of common diseases and variable drug responses. Trends Genet 19:615–622.[ISI][Medline]
Harty LC, Garcia-Closas M, Rothman N, Reid YA, Tucker MA, Hartge P (2000). Collection of buccal cell DNA using treated cards. Cancer Epid Biomarkers Prev 9:501–506.
Hinds DA, Stokowski RP, Patil N, Konvicka K, Kershenobich D, Cox DR, et al. (2004). Matching strategies for genetic association studies in structured populations. Am J Hum Genet 74:317–325.[ISI][Medline]
Hosono S, Faruqi AF, Dean FB, Du Y, Sun Z, Wu X, et al. (2003). Unbiased whole-genome amplification directly from clinical samples. Genome Res 13:954–964.
Lord PG, Papoian T (2004). Genomics and drug toxicity (editorial). Science 306:575.[Abstract]
Meier B (2004). Earlier Merck study indicated risks of Vioxx. New York Times 11/18/2004.
Psaty BM, Furberg CD, Ray WA, Weiss NS (2004). Potential for conflict of interest in the evaluation of suspected adverse drug reactions. Use of cerivastatin and risk of rhabdomyolysis. J Am Med Assoc 292:2647–2650.
Strom BL (2004). Potential for conflict of interest in the evaluation of suspected adverse drug reactions: a counterpoint. J Am Med Assoc 292:2643–2646.
Walker AM (2001). Pattern recognition in health insurance claims databases. Pharmacoepidemiol Drug Safety 10:393–397.[ISI][Medline]
Zavras AI (2004). Using an administrative medical claims database to identify the pharmacogenetics of rare outcomes: scientific, ethical and social issues. In: Pharmacogenomics. Proceedings of the 2004 Cold Spring Harbor Laboratory Conference in Pharmacogenomics, Nov. 18–21, 2004. New York: CSHL, p. 115.
| ||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| IADR Journals | Advances in Dental Research ® |
| Journal of Dental Research ® | Critical Reviews (1990-2004) |
