Cancer incidence during the time period from 1992 to 2004 has been decreasing, due mostly to a favorable trend among men and among those aged older than 65 years. However, there is considerable variation across cancer sites in both the magnitude and the direction of cancer incidence trends over this time period. Here we discuss the likely reasons for variations in incidence trends and speculate on the changes that can be expected between now and the year 2015. We first discuss trends by specific cancer sites and then by cross-cutting issues of tobacco, obesity, screening, and race/ethnicity.
Prostate Cancer. The prostate is the leading site for cancer incidence among US men.1 Prostate cancer incidence has been extremely variable in the past 20 years, largely due to the advent of PSA screening. A sharp increase in incidence began in the late 1980s, almost certainly due to the advent of PSA testing and the detection of a high number of previously undiagnosed, prevalent cases.20 Despite approximately 40% higher rates of prostate cancer incidence among African Americans as compared with Whites, declining trends are similar for both Whites and African Americans. Modifiable nutritional factors may also be important in prostate cancer, especially vitamin E and selenium.21 If the ongoing Selenium and Vitamin E Cancer Prevention Trial (SELECT) finds that either type of supplement can reduce prostate cancer risk, there could be benefits for lower incidence in the coming decade.22 Prostate cancer incidence can be reduced with finasteride chemoprevention, which inhibits the conversion of testosterone to dihydrotestosterone, but this medication has not been commonly used for this purpose because of concern about possible adverse effects leading to cancers with higher Gleason grades.23 Hopefully, findings from PSA screening trials will clarify the value of screening for reducing morbidity and mortality from prostate cancer, as well as the overall impact on duration and quality of life.24,25 The trends of declining prostate cancer incidence will be largely dependent on the rates of PSA testing in the years to come.
Breast Cancer. The breast is the leading site of cancer incidence in US women.1 Over the time period of 1992 to 1999, no substantial changes in overall incidence rates of invasive breast cancer were observed, but after 1999, breast cancer incidence began to decline. Initially the decline was likely due to the saturation of mammography screening, which had already identified many prevalent early breast cancers, ultimately impacting treatment and future mortality trends.18–19 This decline steepened substantially after 2002, primarily due to declines in estrogen receptor-positive tumors among women aged 50 to 69 years.26 This steeper decline since 2002 may be due to the combined effects of a decline in the rate of mammography screening and the sudden decline in HT following the 2002 publication of the Women's Health Initiative for combined estrogen and progestin.13,26,27 Both of these factors will cause a continued decline in breast cancer incidence in the coming years. The obesity epidemic has had adverse effects on breast cancer incidence trends. Without past increases in obesity, incidence declines might have been steeper and seen much earlier. According to the National Health and Nutrition Examination Survey, there has been no significant increase in prevalence of obesity among women between 1999 to 2000 (33.4%) and 2003 to 2004 (33.2%).12 Despite this suggestion that obesity might be stabilizing among women, the current prevalence of overweight and obesity is still quite high (62%).12 Weight gain and excess adiposity are important modifiable risk factors for postmenopausal breast cancer.28 Therefore, if the obesity epidemic can be slowed and reversed in the coming decade, this could cause additional reductions in future breast cancer incidence. Future declines in breast cancer incidence may also be seen as the consequence of removal of atypical hyperplasia and ductal carcinomas in situ that were identified as suspicious lesions by mammography in past years. Tamoxifen and raloxifene have both been shown to substantially reduce the risk of incident breast cancer.29,30 The safety profile for tamoxifen discourages its widespread use, but raloxifene seems to have a better balance of risks and benefits.30 Raloxifene, currently used in the prevention and treatment of osteoporosis, was prescribed in only 12% of patient visits for osteoporosis in 2003.31 If the use of raloxifene increases substantially in the coming years, breast cancer incidence may be expected to fall. In the coming decade, the longer-term effects of decreased use of HT, increased chemoprevention, and slowing of the obesity trends should lead to continued decreases in breast cancer incidence rates.
Lung Cancer. The lung is the second leading site for cancer incidence and the leading site for cancer death among both US men and women.1 Lung cancer incidence rates are approximately 1.7 times higher in men than in women. The downward trend of lung cancer incidence in men is exceeding the 25% reduction goal, but the trends among women are not (see Figure 1). The primary cause of lung cancer is tobacco use, so incidence trends are largely a reflection of tobacco-use trends over the preceding 20-year period.32,33 The prevalence of smoking declined from 52% to 33% among men and from 34% to 28% among women during the time period from 1965 to 1985. Between 1985 and 1995, there was about a 5% decrease in the prevalence of tobacco use among both men and women.34 Despite a persistently higher rate of lung cancer among African Americans than among Whites, a steeper decline has been observed for African Americans compared with Whites over this time period, likely due to historical changes in smoking. Declines in lung cancer incidence have also been observed due to reductions in occupational carcinogen exposures; however, the relative contribution to overall lung cancer rates of these exposures is small compared with tobacco use.32 Screening is not recommended for lung cancer in the general population, but low-dose computed tomography (CT), chest x-ray, sputum cytology, molecular sputum testing, or a combination of these tests are still under investigation and, therefore, may hold promise.24,35, , , , –40 Promising findings for screening with spiral CT scans have led to the implementation of large randomized clinical trials (RCTs) now underway that will be completed by 2010.37 Apart from possible effects of screening, the incidence rates for lung cancer will likely decline in the coming decade as a consequence of past tobacco trends. If CT screening begins to be used widely, then incidence rates will substantially increase as an artifact of the initiation of screening and the detection of prevalent cases, as was observed for prostate cancer in the early 1990s. Apart from this potential artifact of screening, the major factor that will determine lung cancer incidence in the coming decade is the past history of tobacco use. Incidence will, therefore, likely continue to decline among men and soon begin to decline among women.
Colorectal Cancer. Modifiable risk factors associated with the development of colorectal cancer include physical inactivity; adiposity; cigarette smoking; and diets high in red meats, processed meat, or high energy intake, while preventive factors include the use of nonsteroidal anti-inflammatory drugs (NSAIDs); HT; and diets high in fruits, vegetables, calcium, and/or vitamin D.41 Colorectal cancer incidence rates increased until 1985, when they began to decline.42 The reasons for this decline are not clear, but could be tied to downward trends in cigarette smoking, increasing NSAID use, and increasing HT use. The recent decline in HT use may adversely affect colorectal cancer trends among women in the coming years, as HT reduces risk.13 Recent trials have demonstrated the potential for NSAIDs to reduce colorectal adenomas, but adverse effects from these agents will limit their widespread use for that purpose.43,44 Colorectal screening (especially colonoscopy and flexible sigmoidoscopy) leads to the identification and removal of adenomas, thus substantially reducing the risk of incident colorectal cancer.45 Colorectal screening rates (mostly colonoscopy) have been increasing in recent years (Table 3). Colorectal screening is higher among Whites than among African Americans,46 which may help explain the steeper decline observed among Whites. The current rate of decline in the incidence of colorectal cancer is on target to meet the 25% reduction challenge goal for 2015. This favorable trend in colorectal cancer incidence is occurring in spite of the obesity epidemic. If obesity trends can be improved and if we can continue progress in the use of colonoscopy for colorectal screening, the reduction in incidence of colorectal cancer may well exceed the 25% goal.
Uterine Cancer. The primary risk factor likely to affect the current trends in endometrial cancer is obesity.47,48 Obesity likely influences endometrial cancer risk in premenopausal women by causing androgen excess with altered ovarian physiology, whereas in postmenopausal women, adipose tissue increases estrogen production through the conversion of androgens to estrogens by aromatase.49 Another important risk factor for uterine cancer is the use of unopposed estrogens in HT.50 It is likely that the sharp declines in the use of systemic unopposed estrogen8 after 20028 have led to only a small decrease in incidence, as probably only a small number of women who have not had hysterectomies would have been taking unopposed estrogen. However, a recent report from the Women's Health Initiative suggests a lower prevalence of coronary artery calcification among postmenopausal women aged 50 to 59 years who had undergone hysterectomy and received conjugated equine estrogen as compared with those who received placebo.51 These findings, along with earlier findings suggesting that combined HT confers greater postmenopausal breast cancer risk13 than estrogen alone,14 may lead to an increase in the use of unopposed estrogen therapy in women with uteri, thus increasing their risk of endometrial cancer. Surveillance of specific prescribing factors for HT should be monitored. Overall, the high obesity prevalence may influence future trends. Therefore, the most reasonable prediction would be a continued stable rate of uterine cancer.
Bladder Cancer. Bladder cancer is the fourth leading cancer among US men.1 The primary modifiable risk factors for bladder cancer are cigarette smoking and occupational exposures to carcinogens.52 However, the anticipated decline in incidence of bladder cancer due to past reductions in tobacco use and occupational exposures has not been seen over the time period of 1992 to 2004. The reasons why bladder cancer rates have remained unchanged are not clear, hence it is not possible to confidently predict changes in the coming decade.
Non-Hodgkin Lymphoma. Very little is known about modifiable risk factors for non-Hodgkin lymphoma other than HIV/AIDS, Epstein-Barr virus, herpes virus 8, human T-cell lymphotropic virus, and immunosuppressive drugs.53 Potential roles of other infectious agents and occupational and environmental factors such as benzene54 or polychlorinated biphenyls,55 as well as others,56 are unclear. The most reasonable prediction for the coming decade is, therefore, continuation of the past trend of an increase in rates.
Melanoma. Melanoma rates have been increasing substantially in recent years. This is likely due to the combined effects of previous sun exposures and increased diagnosis of very small cancers due to improved awareness and surveillance of pigmented lesions.57 There has been a modest 8% increase in sunscreen use among adolescents from 1998 to 2004,58 but no significant decrease in reported sunburn in the previous summer (69% versus 72%).46 As most melanoma occurs in older people, it is likely rates will continue to increase into the coming decade resulting from past sun exposures.
Ovarian Cancer. The few modifiable risk factors that decrease risk include oral contraceptives, hysterectomy, tubal ligation, and high parity.59 In 1982, only 76% of US women had ever used the pill as compared with 82% in 1995 and 2002.60 Past trends in oral contraceptives may account for the observed declines in ovarian cancer incidence. There are currently no recommendations for screening in the general population, but 2 large screening trials are ongoing.61 Because the impact of oral contraceptive use on ovarian cancer risk persists for at least 15 years,62 recent trends in their use will likely lead to continued declines in ovarian cancer incidence in the coming decade.
Leukemia. Leukemia is diagnosed 10 times more frequently in adults, but it is the leading cause of cancer in children aged 0 to 14 years.63 Leukemia is comprised of several diverse types, with few having any identifiable modifiable risk factors.64 Nevertheless, occupational exposures, radiation exposure, chemotherapy, and smoking have all been implicated as risk factors for acute myeloid leukemia, the most frequently diagnosed leukemia.64 It is most reasonable, therefore, to predict that the unchanging rates will continue into the coming decade.
Cancers of the Oral Cavity and Pharynx. The primary modifiable risk factor for oral cancer is tobacco exposure, either by smoking or chewing.65,66 Alcohol is also a factor that works synergistically with tobacco.67 The decline in oral cancer between 1992 and 2004 is likely due to historical reductions in tobacco exposure. Continuing declines are likely in the coming decade.
Pancreatic Cancer. Cigarette smoking is the major risk factor for pancreatic cancers.66,68,69 The importance of other risk factors, including obesity and vegetable intake, is less certain. Despite declining cigarette exposures over the past several decades, rates of pancreatic cancer have remained stable.70 This may be due in part to the more certain diagnosis of pancreatic cancer due to improved diagnostic imaging. Based on the past trends, no substantial change is likely in pancreatic cancer rates in the coming decade.
Kidney Cancer. Increased incidence of kidney cancer has been primarily attributed to small localized tumors identified in patients who undergo diagnostic evaluation for unrelated conditions.71 However, this may not completely explain the increased trend because other factors may also be contributing, including obesity trends,72 as obesity is an important risk factor.47,48 This rising trend in incidence has occurred in spite of decreasing prevalence of tobacco use, a major risk factor for kidney cancer.66 It is expected that the incidence of kidney cancer will continue to rise in the coming decade.
Stomach Cancer. The incidence of stomach cancer has been declining over the past several decades in the United States.73 Declining rates in the historical trends of stomach cancer have been attributed to the nutritional benefits coming from improved food storage and distribution systems and to the declining prevalence of smoking and chronic infection with Helicobacter pylori.74,75 It is likely the long-term historical decline in stomach cancer will continue into the coming decade due in part to declines in smoking prevalence and to declines in the prevalence of persistent Helicobacterpylori infection initiated in childhood among younger cohorts.
Myeloma. Because the etiologic factors for myeloma are not well understood,76 it is most reasonable to project the future rates will remain stable.
Liver Cancer. Liver cancer incidence has been substantially increasing in the past decades.77, –79 The primary risk factors for liver cancer include chronic infection with hepatitis B virus and hepatitis C virus, along with excess alcohol and obesity.77, –79 The risk of liver cancer is much higher among foreign-born persons, particularly Asian/Pacific Islanders, due to their high rates of chronic infection with hepatitis B virus.80 Despite the initiation of universal infant/childhood hepatitis B vaccination programs for children in the past 20 years,81 only limited effects on hepatocellular carcinoma will be observed by 2015. Based on immigration trends and the epidemic of hepatitis C virus infection, prediction models suggest that liver cancer will continue to rise in the United States over the coming 15 years.82,83
Thyroid Cancer. The incidence of thyroid cancer has been increasing in the United States for the past several decades, primarily due to an increase in small papillary cancers.84 While the increase in small papillary cancers is most likely due to increased detection from improved medical imaging and diagnostic techniques, it is uncertain whether there are any other reasons contributing to this trend.84,85 With increasing numbers of people undergoing neck ultrasound examinations, this trend will likely continue into the future.
Cancers of the Brain and Nervous System. The incidence of brain cancer increased in the 1980s, perhaps due to the advent of better imaging methods for the brain, but rates have been steady in recent years. As little is known about the etiologic factors for brain cancer, it is not possible to confidently predict the trends in the coming decade.86
Cervical Cancer. Invasive cervical cancer is uncommon in the United States because of widespread screening using Pap smears that identify and remove precursor lesions.87 In 2006, the Food and Drug Administration approved a human papillomavirus vaccine for use in girls and women aged 9 to 26 years.88 This vaccine has been shown to be highly effective in protecting against the human papillomavirus serotypes that together cause about 70% of cervical cancer cases.89 Little effect on cervical cancer incidence due to the use of this vaccine will be observed before 2015. Nonetheless, given the historical declining trend and continued high prevalence of Pap smear screening,7 the United States has already surpassed the 25% reduction goal for cervical cancer, and incidence will likely continue to decline, but at a slower rate in the future, as no substantial changes in the prevalence of Pap smear screening are likely.
Esophageal Cancer. The overall incidence of esophageal cancer has remained fairly constant over the past 12 years. However, trends in incidence have been decreasing for the more common squamous cell carcinoma of the esophagus and increasing for adenocarcinoma of the esophagus.90,91 Declines in the rates of squamous cell carcinoma of the esophagus are likely due to declining rates of smoking and alcohol consumption, while increases in adenocarcinaoma of the esophagus are likely due to factors causing acid-reflux disorders of the lower esophagus, especially abdominal obesity.90,91 Due to the trends in tobacco and obesity, it is expected that the past trends of reductions in the more common squamous cell cancer will continue, and increases in adenocarcinoma of the esophagus will also likely continue into the coming decade.