The Public Health Impact of DCEG Research

February 19, 2013, by Shelia Hoar Zahm, Sc.D.

Many DCEG research findings have served as key evidence by expert panels that develop consensus reports as to whether specific exposures cause cancer, by regulatory agencies that establish permissible exposure levels, and by the medical community to change clinical practice. Over the years, DCEG has had a significant impact on public health through the use of its research by national and international organizations, including the International Agency for Research on Cancer (IARC), the National Toxicology Program’s Report on Carcinogens (ROC), the Environmental Protection Agency (EPA), the Food and Drug Administration (FDA), and the Occupational Safety and Health Administration (OSHA); various agencies dealing with radiation safety; and many professional societies. Selected landmark findings include the following:

Smokeless Tobacco
Diet/Nutrition/Physical Activity
Radiation
Pharmaceuticals and Medical Devices
Ultraviolet Radiation
Infectious Agents
Electromagnetic Fields
Hereditary Syndromes
Occupation
Breast Cancer
Indoor Air Pollution
Melanoma
Water Contaminants

Smokeless Tobacco

The U.S. Cancer Mortality Atlas revealed a striking elevation of oral cancer rates among women in the rural south. In a case-control study conducted in North Carolina, an elevated risk was associated with long-term use of snuff among nonsmokers, reaching 50-fold for tissues in direct contact with tobacco. The published report in the New England Journal of Medicine (Winn et al., 1981) stimulated congressional hearings that resulted in regulatory actions to control the advertising and labeling of smokeless tobacco and educational campaigns to reduce its use by young people.

Radiation

Studies of cancer and leukemia among atomic bomb survivors and a pooled analysis of radiation-related risk of thyroid cancer from various exposures have greatly influenced national and international radiation safety standards (Ron et al., 1994, 1995, 1998; Land et al., 1995, 1996, 2003).
DCEG developed statistical models to calculate the radiation-related probability of cancer causation, known as the NIH radioepidemiological tables (NIH, [1985 pdf], [2003 pdf]), including an online interactive program (Kocher et al., 2008) used by the National Institute for Occupational Safety and Health and the Veterans Administration for adjudicating compensation claims.
Reports of the potential health risks of exposure to radioactive I-131 fallout from above-ground nuclear testing in the 1950s and 1960s formed the basis of national educational campaigns (NCI, 1997; Gilbert et al., 1998).
The quantification of lung cancer risks associated with occupational and residential exposures to radon were critical to informing the EPA action level for radon mitigation (Qiao et al., 1989; Lubin et al., 1990, 1997, 1998, 2004; Alavanja et al., 1994, 1999; Wang et al., 2002).
The study of second primary tumors after childhood cancer demonstrated a radiation dose–related relationship for certain cancers, such as bone sarcoma and thyroid cancer (Tucker et al., 1987, 1991), which has led to efforts to decrease doses of radiotherapy in children.
The striking dose-related risk of radiation-related second cancers among children treated for retinoblastoma (Kleinerman et al., 2005) has altered medical practice to reduce radiation exposure and intensify second cancer screening.
Quantification of the dose response for lung and breast cancers after radiation treatment of Hodgkin lymphoma (Tucker et al., 1988; Travis et al., 2002, 2003) influenced clinical standards to reduce radiation fields, lower exposures, and avoid smoking.
The elevated risk of radiation-related solid cancers reported among bone marrow transplant recipients affected clinical practice by moving the conditioning regimens away from radiation and using chemotherapy-based regimens as an alternative (Curtis et al., 1997; Socie et al., 2000; Rizzo et al., 2009).
Statistical models estimating the cancer risks associated with radiation exposure from computed tomography (CT) scans (Berrington de González et al., 2009) led to a requirement that makers of scanners used at NIH incorporate software to measure and track patients’ radiation doses over time, an improvement that is being adopted by other hospitals and imaging facilities.
A professional education brochure was created in collaboration with the Society for Pediatric Radiology in efforts to minimize the level of radiation from CT scans to children (2002, 2009).
Concern was heightened by a recent epidemiological finding of small but statistically increased risks of leukemia and brain tumors in the first decade after CT exposure in children (Pearce et al., 2012).

Ultraviolet Radiation

Surveys of ground-level ultraviolet radiation in relation to time trends for melanoma and other skin cancers have provided EPA with the estimated prevalence of skin cancer in relation to projected increases in ozone depletion (Scotto et al., 1975, 1981; Fears et al., 1976, 1977, 1983).

Electromagnetic Fields

The finding that cell phone use did not increase risk of brain tumors affected FDA’s regulatory policies on radiofrequency radiation (Inskip et al., 2001).
Lack of an increased risk for childhood leukemia associated with high residential magnetic fields formed the basis of judicial rulings in the United Kingdom and elsewhere on the potential health effects of these exposures (Linet et al., 1997).

Occupation

The excess risk of respiratory cancer among copper smelter workers (Lee and Fraumeni, 1969) and nearby residents (Blot and Fraumeni, 1975) exposed to inorganic arsenic has led to IARC’s classification of airborne arsenic as a human carcinogen and to new OSHA exposure limits.
The increased occupational risks of bladder cancer (Silverman et al., 1983) and lung cancer (Silverman et al., 2012) led IARC to classify diesel exhaust fumes as an established human carcinogen.
Elevated risks of nasopharyngeal cancer and leukemia among formaldehyde-exposed workers (Blair et al., 1986; Hauptmann et al., 2003, 2004, 2009; Beane Freeman et al., 2009) were key to IARC’s and ROC’s classifications of formaldehyde as a human carcinogen and led to EPA regulatory standards.
Evidence of hematotoxicity from benzene exposure at levels below the occupational standard at the time (Lan et al., 2004) led to a lower action level in China and informed an EPA rule limiting the benzene content in gasoline and requiring controls on passenger vehicles and portable fuel containers to reduce pollutants.
Studies of silica-exposed miners and workers in pottery factories and dusty trades (Chen et al., 1989, 1990, 1992; Amandus et al., 1991, 1995) were cited in IARC’s classification of silica as a human carcinogen.
Excess lung cancer reported among chromium pigment workers (Hayes et al., 1989) was among the study findings that led to IARC’s classification of chromium compounds as carcinogenic.
Studies showing increased nasal cancer among furniture workers (Brinton et al., 1977, 1984) influenced IARC’s classification of wood dust exposure as a human carcinogen.
A report of increased lymphoma risk associated with 2,4-D exposure (Hoar et al., 1986) led to EPA label instructions to change clothes after application and influenced the Veterans Administration’s compensation policies for exposure to Agent Orange.

Indoor Air Pollution

Increased lung cancer associated with smoky coal use (Lan et al., 2002, 2008) and cooking fumes (Gao et al., 1987) in China were key to IARC’s classification of indoor emissions from household combustion of coal as an established human carcinogen and emissions from high-temperature frying as a probable human carcinogen. The findings also informed the World Health Organization’s WHO Guidelines for Indoor Air Quality: Selected Pollutants.

Water Contaminants

A study linking disinfection byproducts to bladder cancer risk (Cantor et al., 1987) stimulated EPA to lower the allowable maximum contaminant levels for trihalomethanes and haloacetic acids in drinking water.
No epidemiologic evidence was found that fluoride in drinking water poses an elevated risk of cancer, as had been suggested by some previous reports. The finding was confirmed by expert panels convened in the United States, the United Kingdom, and other countries, providing further confidence in the safety of water fluoridation (Hoover et al., 1976).

Diet/Nutrition/Physical Activity

The lack of an association between saccharin use and bladder cancer risk (Hoover and Strasser, 1980) was part of the evidence that prompted ROC to delist saccharin as a potential carcinogen.
Several studies indicating that diets high in vegetables and fruits are associated with reduced risk of various cancers (Ziegler et al., 1981, 1986; Winn et al., 1984) influenced public health recommendations.
The Nutrition Intervention Trials in Linxian, China, were the first randomized trials to demonstrate that vitamin/mineral supplementation reduces total and cancer mortality in a population with nutritional deficiencies (Blot et al., 1993).
Studies of serum levels of vitamin D in relation to cancer mortality (Freedman et al., 2007) and breast cancer risk (Freedman et al., 2008) were key to the Institute of Medicine’s 2010 report Dietary Reference Intakes for Calcium and Vitamin D.
Studies reporting an increased risk of pancreatic cancer associated with vitamin D intake (Stolzenberg- Solomon et al., 2006, 2009, 2010) have dampened enthusiasm for supplementing vitamin D above a physiologic dose.
Reports that carcinogenic heterocyclic amines are generated in meat during high-temperature cooking (Sinha et al., 1994, 1995) led to public health recommendations on cooking practices.
Studies indicating a reduced risk of colon, breast, endometrial, and ovarian cancers (Albanes et al., 1989; Chow et al., 1993; Dosemeci et al., 1993; Sturgeon et al., 1993; Zheng et al., 1993) were cited in the 1996 Surgeon General’s Report on Physical Activity and Health.

Pharmaceuticals and Medical Devices

Long-term follow-up of cohorts exposed prenatally to diethylstilbestrol (DES) revealed a broader spectrum of adverse health outcomes than previously known. These findings have affected medical surveillance of exposed persons and have sparked research on the effects of early-life chemical exposures beyond DES (Hatch et al., 1998, 2001, 2011; Titus-Ernstoff et al., 2001, 2010; Troisi et al., 2007a, 2007b; Hoover et al., 2011).
The finding that hormone replacement therapy is associated with increased breast cancer risk—and that the risk is greater with estrogen-progestin regimens than with estrogen alone (Hoover et al., 1976; Schairer et al., 2000)—has changed medical practice, resulting in a decrease in breast cancer incidence.
Evidence against breast implants causing an increase in breast cancer or connective tissue disorders (Brinton, et al., 2004, 2006) contributed to FDA’s decision to allow the devices back on the market.
Evidence that ovulation-stimulating agents do not significantly increase the risk of ovarian and other cancers (Brinton, Lamb et al., 2004; Brinton, Scoccia et al., 2004) has reduced concern about prescribing these increasingly popular drugs.
A study of leukemia after childhood cancer revealed a dose-response relationship with alkylating agent chemotherapy, after adjusting for radiation (Tucker et al., 1987), which led to decreased use of these agents for the treatment of childhood cancer.
A dose-response risk of childhood leukemia was seen following treatment with chloramphenicol (Shu et al., 1987). As a result of this and other evidence, chloramphenicol is no longer the first-line agent for any infection in developed countries.
A study of secondary leukemia after treatment of breast cancer affected clinical practice by demonstrating the substantial risk associated with melphalan and the relative safety of the current cyclophosphamide-based chemotherapy regimens (Curtis et al., 1992).
Quantification of the increased risks of lymphoma and certain other cancers associated with immunosuppressive drugs in transplant recipients (Hoover and Fraumeni, 1973) led to FDA warnings as well as screening recommendations.
Evidence showing an excess risk of leukemia among women treated with chlorambucil for ovarian cancer (Greene et al., 1982) was key to IARC’s classification of the drug as a human carcinogen and to FDA clinical alerts.

Infectious Agents

A study of newborns who were inadvertently given SV40-contaminated polio vaccine suggested no increased risk of cancer (Fraumeni et al., 1963; Fraumeni et al., 1970; Mortimer et al., 1981), thereby alleviating public health concerns about exposure to this oncogenic virus.
The increased risks of human T-cell leukemia virus type 1 (HTLV-I) and HTLV-I–associated myelopathy/ tropical spastic paraparesis (HAM/TSP) found among persons infected with HTLV-I (Blattner et al., 1982; Blayney et al., 1983) were critical to FDA recommendations to screen blood donations for HTLV-I/II.
DCEG research provided the initial assessment of the specificity, sensitivity, and appropriate applications of the first-generation HIV antibody testing system for diagnosis of HIV infection (Weiss et al., 1985).
A prospective study of HIV infection and the development of AIDS in subjects with hemophilia showed that a much larger proportion of HIV-infected persons would develop AIDS than previously thought (Goedert et al., 1989), which had public health and clinical impact.
Recognition that CD4 count and HIV load predict risk for AIDS and death (Goedert et al., 1987, 1989; Ehmann et al., 1994; O’Brien et al., 1996) had clinical applications for screening and counseling.
A study of HIV infection among laboratory workers influenced new recommendations for handling HIV in the laboratory (Weiss et al., 1988).
New “back calculation” statistical methods were used to derive estimates of HIV prevalence by the U.S. Public Health Service and showed that the HIV prevalence was greatest among persons at risk through heterosexual contact (Rosenberg et al., 1991; Rosenberg, 1995; Rosenberg and Biggar, 1998).
Demonstration that serological tests for human herpesvirus 8 had poor reproducibility (Rabkin et al., 1998) contributed to FDA’s decision not to screen the U.S. blood supply.
Landmark studies of the natural history of cervical cancer firmly established human papillomavirus (HPV) as the necessary cause (Schiffman et al., 1993, 2007), laying the groundwork for vaccine and screening strategies.
The finding that one or two doses of an HPV 16/18 vaccine may prevent cervical cancer just as effectively as three doses is likely to lower the cost and encourage the global dissemination of the vaccine (Kreimer et al., 2011).
Studies into the benefits of incorporating HPV testing into cervical cancer screening programs (Katki et al., 2011) have informed screening guidelines issued by the U.S. Preventive Services Task Force, the American Cancer Society, the American Society for Clinical Pathology, and the American Society for Colposcopy and Cervical Pathology.
Studies linking Helicobacter pylori infection to the very high incidence of gastric cancer and precancerous lesions in Shandong province in China prompted a clinical trial that showed a sustained reduction in risk following a two-week course of antibiotics (You et al., 2006; Zhang et al., 2006; Ma et al., 2012).

Hereditary Syndromes

Identification of the inherited genes responsible for hereditary cancer syndromes and mutation carriers has had a significant impact on the clinical management of these and related conditions. Examples include the discovery of Li-Fraumeni syndrome and the role of p53 (Li and Fraumeni, 1969, 1982; Li et al., 1988; Malkin et al., 1990) and the roles of NF2 in neurofibromatosis type 2 (Rouleau et al., 1993; Trofatter et al., 1993); CDKN2A, CDK4, and MITF in hereditary melanoma (Hussusian et al., 1994; Zuo et al., 1996; Yokoyama et al., 2011); PTCH in nevoid basal cell carcinoma syndrome (Hahn et al., 1996); SUFU in medulloblastoma (Taylor et al., 2002); and T (brachyury) duplication in familial chordoma (Yang et al., 2009).
Detailed clinical studies of neurofi-bromatosis type 2 revealed heterogeneity of the disease phenotype, changed clinical management, and informed genetic counseling (Kaiser- Kupfer et al., 1989; Parry et al., 1994, 1996; Ruttledge et al., 1996).
Studies of monoclonal B-cell lymphomatosis as a precursor condition for chronic lymphocytic leukemia (CLL) in high-risk families and in the general population have facilitated screening and early diagnosis of CLL (Landgren et al., 2009; Goldin et al., 2010).
The observations that patients with dyskeratosis congenita (DC) have extremely short telomeres and that approximately 60 percent have a germline mutation in a telomere biology gene led to the development of telomere length as a diagnostic test for DC and new criteria for evaluating potential bone marrow donors (Alter et al., 2007; Savage et al., 2008).
A novel chromosome 9q deletion in a patient with nevoid basal cell carcinoma syndrome followed by genetic linkage and fine-mapping studies led to the identification of the PTCH gene as the cause of this syndrome (Gailani et al., 1992; Hahn et al., 1996; Chidambaram et al., 1996). This work set in motion a research effort that recently culminated in the first FDA-approved biological agent (vismodegib) that targets the Hedgehog signaling pathway, a novel therapy for locally advanced and metastatic basal cell carcinoma of the skin.
Publication of the Concise Handbook of Familial Cancer Susceptibility Syndromes has provided a useful reference for clinical recognition and management of these rare but important disorders (Lindor and Greene, 1998; Lindor et al., 2008).

Breast Cancer

The Gail breast cancer risk assessment model (Gail et al., 1989) is widely used in risk prediction strategies, including FDA guidelines on the use of tamoxifen and raloxifene for breast cancer risk reduction.
A study of BRCA1 and BRCA2 alterations in the Ashkenazi Jewish population of Washington, D.C., reported lower penetrance for breast and ovarian cancers than earlier studies in families with hereditary breast cancer. The findings highlighted the importance of population-specific founder mutations in genetic testing and affected the clinical management of carriers (Struewing et al., 1997).
Studies in hereditary breast/ovarian cancer uncovered the value of risk-reducing salpingo-oophorectomy, now considered a standard-of-care treatment option in the management of women in high-risk families (Tobacman et al., 1982; Struewing et al., 1995; Kramer et al., 2005; Greene et al., 2008, 2011).
Epidemiological studies using tumor tissue microarrays have revealed the heterogeneity of breast cancer in terms of risk factors and prognosis (García- Closas et al., 2006, 2008; Yang et al., 2007), which has enhanced clinical care decisions.

Melanoma

Identification and characterization of dysplastic nevi as the major risk factor for melanoma informed the guidelines for melanoma screening and clinical management in high-risk families and in the general population (Reimer et al., 1978; Greene et al., 1985a, 1985b; Tucker et al., 1997).
A clinical atlas depicting melanoma and precursor lesions, training videos to help health care providers examine high-risk families, and a risk calculator to estimate an individual’s probability of developing melanoma have been important milestones in managing this disease (Tucker et al., 2002; Fears et al., 2006).