Preventive Oncology 

In many respects, cancer is a preventable disease. Estimates indicate that approximately one half of all cancer cases either arise from modifiable risk factors or can be detected as precursor lesions before the development of disease with metastatic potential.^[1]

Prevention of cancer can take place on several different levels: primary prevention addresses the cause of cancer so disease does not occur, secondary prevention identifies disease before the onset of symptoms and keeps it from becoming more extensive, and tertiary prevention reduces complications and progression of disease once it has become clinically apparent.

Although cancer has overtaken cardiovascular disease to become the leading cause of death among men and women younger than 85 years in the United States, and the number of cancer deaths continues to increase with the aging and growth of the population, age-standardized US death rates from cancer have been decreasing.^[2] From 2001 to 2006, cancer mortality decreased over 1.5% per year. Overall cancer incidence has also decreased, by an average of 0.7% per year from 1999 to 2006.^[3] These declines have been attributed to risk reduction strategies, detection of early disease, and improvement in treatment strategies. In this review, we will address the first 2 of these factors to summarize the evidence for prevention in oncology.

Assessment of an individual's risk is a key step in cancer prevention; risk assessment programs have been developed at many cancer centers to identify people who are at high risk. Review of personal and family medical history, work history, and lifestyle can help identify cancer risk factors, which may be modifiable (eg, tobacco use, sun exposure) or nonmodifiable (eg, family history of cancer, sex, ethnicity, race, advancing age, hormone levels). Combinations of modifiable and nonmodifiable risk factors place some people at particularly high risk for cancer.

Models of cancer risk have been developed to permit calculation of an individual's risk for a specific type of cancer. One of the best known of these is the Gail model, which predicts breast cancer risk on the basis of current age, race, age when menstruation began, age of first live birth, number of close relatives with breast cancer, number of breast biopsies, and the presence or absence of atypical hyperplasia on breast biopsies.^[4] Like most cancer risk models, the Gail model has limitations: it does not include ovarian cancer history or breast cancer in second-degree relatives such as aunts, cousins, or grandparents. This model may also be less accurate in predicting risk in non-white women. Therefore, risk models should be selected based on each individual's situation in order to calculate risk as accurately as possible.

Overall, approximately 10% of cancers occur because of hereditary predisposition, such as mutations in cancer susceptibility genes (eg, BHCA1 and BRCA2). Many of these mutations can be identified through analysis of a blood or tissue sample. Currently, genetic testing is recommended only for individuals who have a personal or family history that is suggestive of an inherited cancer syndrome.

A study by Reulen et al found that survivors of childhood cancer are at excess risk of developing primary neoplasms later in life, with the greatest risk for digestive and genitourinary neoplasms in survivors older than 40 years.^[5]

Identification of the appropriate person for testing leads to more informative results for the entire family. In most cases, the person who has had the cancer that best fits the hereditary pattern should be chosen for testing. If a mutation is identified in this individual, others in the family who are at risk can then be tested for that specific mutation. Genetic analysis should always be preceded by careful genetic counseling, which continues after determination of gene mutation status. Identification of a genetic alteration may change recommendations for cancer screening, chemoprevention, and prophylactic surgery.

In addition to risk from genetic syndromes, it is estimated that approximately another 15-20% of cancers are familial, which may be due to low-penetrance genetic changes or the effects of shared environment and behaviors. The remainder of cancers in the general population are considered sporadic.

Modifiable risk factors for cancer include tobacco use, sun exposure, diet, exercise, obesity, alcohol use, hormone replacement therapy (HRT), environmental/occupational exposures, infectious exposures, and sexual activity.

Tobacco use

Tobacco use accounts for approximately one third of cancer deaths in the US.^[6] Of these, lung cancer is the most common, but cancers of the blood, head and neck, esophagus, liver, pancreas, liver, stomach, cervix, kidney, colon, and bladder have also been linked to smoking.^[7] Many chemicals are present in tobacco smoke, including at least 69 known carcinogens.^[8] Smokeless tobacco—chewing tobacco and dipping snuff—contains at least 28 carcinogens.^[9]

Smoking may also promote more aggressive forms of cancer: for example, tobacco use is associated with higher-grade and higher-stage prostate cancer.^{[10, 11]} Secondhand smoke, also known as environmental tobacco smoke, has been associated with both lung and sinus cancers in nonsmokers.^[12] ln addition, considerable evidence indicates that smokeless tobacco and cigars also have deadly consequences, including lung, laryngeal, esophageal, and oral cancers.^[9]

Carcinogenesis from tobacco use occurs through several mechanisms, including direct delivery of carcinogens to tissues, inflammation, and breakdown of physiologic barriers.^[13] Cessation of tobacco use has been shown to reduce both cancer-related and all-cause mortality. Health benefits start soon after quitting and can be seen even in long-time users.^[14] For former smokers who have been abstinent for 10 years, the risk of lung cancer is one half that of current smokers. This risk falls to as low as 10% for ex-smokers who have quit for 30 years or more. Moreover, the risk for cancers of the mouth, throat, and esophagus lessens significantly 5 years after quitting, and the risk of developing bladder or cervical cancers also decreases after just a few years of being nicotine free.^[15]

Most individuals require several quit attempts before they are able to stop smoking. The addiction to tobacco use is both psychologic and biochemical. Medications are available to address the biochemical aspects, whereas counseling and other social support are recommended for treating the psychologic aspects.

Nicotine replacement therapy provides nicotine without the other components of tobacco and can be administered through a patch, nasal spray, chewing gum, lozenge, or inhaler. Clinical trials have demonstrated that the use of nicotine replacement doubles the rate of quitting tobacco use.^{[16, 17]}

The addition of bupropion (Zyban), an antidepressant medication, may increase the efficacy of nicotine replacement.^[18] A meta-analysis has also shown that bupropion doubles the quit rate when used as monotherapy.^[19] However, this medication should be avoided in individuals with a history of seizures or eating disorders.

The newest medication for smoking cessation is varenicline (Chantix),which was found to be superior to bupropion in 3 clinical trials.^{[20, 21, 22]} Randomized, double-blind studies comparing varenicline and nicotine replacement have not been performed to date, but unblinded studies suggest that the quit rates may be higher with varenicline as well.^{[23, 24]} Individuals with depression or other neuropsychiatric disorders should not take varenicline, as this medication may exacerbate their symptoms.

Of course, the ideal way to reduce tobacco-associated cancers is for people to not start using tobacco in the first place. Community programs for youth and public service announcements are 2 means of addressing this issue.

Sun exposure

Nonmelanoma skin cancers comprise 40% of malignancies in the US. The incidence rates of melanoma, although much less common than nonmelanoma skin cancers, is increasing and this condition has a much higher propensity for metastasis and death.

Ultraviolet radiation is a well-established carcinogen for both melanoma and nonmelanoma skin cancers. However, the patterns of sunlight exposure associated with these cancers differ significantly. Squamous cell carcinoma tends to occur in those who have chronic sun exposure, often due to occupational exposure from working outdoors.^{[25, 26]}

Episodes of intense ultraviolet exposure, particularly in children and others without a history of chronic sun exposure, is associated with melanoma incidence,^[27] and a history of blistering sunburns more than doubles the risk of melanoma.^[28] Basal cell carcinoma, although usually associated with chronic sun exposure, has also been linked to intermittent exposure in a significant proportion of cases.^[29] In addition, the use of tanning beds has been associated with increased skin cancer incidence^[30] ; in fact, the World Health Organization (WHO) has recommended against the use of these devices for individuals younger than age 18 years.^[31]

Recent interest in vitamin D as a chemopreventive agent for several types of cancer has made sun avoidance a controversial topic, because this nutrient is most easily and effectively obtained from sun exposure. Several clinical trials with vitamin D are planned or under way, and this remains an area of active investigation in cancer prevention. However, at this time, most groups recommend limiting exposure to the sun during peak hours (between 10 AM and 4 PM); using protective clothing, including hats and sunglasses; and using sunscreens with a sun protection factor (SPF) of 30 or greater. Sunscreens should be broad-spectrum and contain agents that work against both ultraviolet (UV) A and UVB radiation (eg, oxybenzone, avobenzone, titanium dioxide, or zinc oxide).

Proper use of sunscreen is also critical: regardless of the SPF, all sunscreen should be applied approximately 30 minutes before sun exposure and then reapplied liberally at least every 1.5-2 hours or after swimming or perspiring heavily. A randomized clinical trial demonstrated statistically significant protection against squamous cell carcinoma with regular use of sunscreen; however, it should be noted that this protective effect did not extend to basal cell carcinoma and that the study was underpowered to assess effect on melanoma risk.^[32] Several meta-analyses have assessed the effects of sunscreen use on melanoma risk and found no association.^{[33, 34, 35]} Therefore, sun avoidance and clothing protection remain the mainstays of skin cancer prevention as well.

Diet

Multiple components of the diet have been studied, both singly and in combination, for their effects on cancer risk. However, study of diet has proven difficult. Epidemiologic studies may have limited reliability because of inaccurate recall and multiple confounding factors, whereas clinical trials are subject to noncompliance, inappropriate form of the nutrient, or an insufficient follow-up period.^[36]

Total dietary fat appears to affect the incidence of prostate cancer but has not been consistently proven to affect rates of colon or breast cancer.^[37] lt should be noted that fat intake is generally evaluated with adjustments for caloric intake and weight gain in order to isolate any effects of the fats themselves. Additional investigation of the effects of types of fat is ongoing. Individuals with high levels of consumption of red meat have been found to have an increased risk of colorectal cancer.^{[38, 39, 40, 41]}

Some dietary practices appear to reduce cancer risk. The link between fruit and vegetable intake and decreased overall cancer risk is weak^{[42, 43]} ; however, pooled analyses suggest a 25% reduction in the incidence of distal (but not proximal) colon cancer in those who consumed more than 800 g of fruits and vegetables per day.^[44] Studies focused on tomato products suggest a benefit for prostate cancer risk, possibly due to the lycopene content in these foods.^[45]

Consumption of calcium has been associated with a decreased incidence of any cancer in women and a decreased incidence of colon cancers in both men and women.^[46] Epidemiologic studies have found that increased fiber intake is associated with a decreased risk of colon cancers and adenomas; however, clinical trials of a high-fiber diet showed no effect on the rates of adenoma recurrence.^{[47, 48, 49, 50]}

Overall, dietary recommendations from the American Cancer Society (ACS) include (1) eating a variety of healthful foods, with 5 or more servings of vegetables and fruits per day; (2) use of whole grains in preference to processed (refined) grains and sugars; and (3) limited consumption of red meats, especially processed meats and those high in fat.^[51]

Exercise

Estimates indicate that a sedentary lifestyle is responsible for approximately 5% of cancer deaths.^[52] Higher levels of physical activity have been associated with decreases in the risks of colon and breast cancers, and possible decreased risks of endometrial, prostate, liver, pancreatic, stomach, and lung cancers have been described as well.^{[53, 54, 55, 56, 57, 58, 59]} For colon and breast cancers, the benefit of physical activity has been demonstrated at multiple weight levels, which implies that the effect of the activity is independent of an effect of weight reduction. The mechanism of the protective effect remains uncertain but may be related to effects on immunity, hormone levels, or prostaglandins.

The frequency, duration, or intensity of exercise needed to prevent cancer has not been determined definitively. The American Cancer Society recommends that adults engage in moderate exercise for at least 30 minutes on 5 or more days per week. Increasing the length of exercise to 45 minutes or more may provide additional protection against breast and colon cancer. For children, the recommendation is to exercise for at least 60 minutes per day on at least 5 days per week.^[51]

Obesity

Epidemiologic studies have indicated that excess weight or obesity result in 14% of cancer deaths in men and 20% of cancer deaths in women.^[60] The cancers associated with obesity are similar to those associated with decreased physical activity. Excess weight has been found to account for 10-40% of colorectal, endometrial, renal, esophageal, and postmenopausal breast cancers.^[61]

Possible links have also been described for some hematologic malignancies and cancers of the prostate, liver, gallbladder, pancreas, stomach, ovary, and cervix, as well. Interventions such as bariatric surgery may reduce the risk of cancer deaths by as much as 60%.^[62]

Alcohol use

Long-term alcohol use has been associated with approximately 4% of incident cancer cases.^[63] In a study of a large database of women, 1 alcoholic beverage per day increased overall cancer risk by 6%.^[64] The increase in cancer incidence in this group involved increased risks of cancers of the head and neck, esophagus, rectum, liver, and breast. The Women's Health Study also showed an increase in breast cancer risk associated with moderate alcohol consumption.^[65]

The mechanism of carcinogenesis associated with alcohol use is not fully understood at this time but may involve inflammatory, epigenetic, hormonal, or metabolic effects. Several of the metabolites of ethanol have been identified as carcinogens.^[66] The American Cancer Society recommends limiting alcohol use to 2 drinks per day for men and 1 drink per day for women (due to the slower metabolism of alcohol by women). A drink is 12 ounces (oz) of beer, 5 oz of wine, or 1.5 oz of 80-proof liquor.

Hormone replacement therapy

Considerable epidemiologic evidence suggests that the duration of a woman's exposure to endogenous estrogen affects breast cancer risk. Support for this hypothesis includes the increased risk of breast cancer for women with an earlier age of menarche, later age of menopause, nulliparity, and later age at first live birth, as well as higher serum estrogen concentrations. For women who take estrogen-­only hormone replacement therapy, meta-analyses of epidemiologic studies indicate a mildly increased risk of breast cancer.^{[67, 68, 69]}

The primary clinical trial to investigate this question was the Women's Health Initiative trial, in which estrogen-only hormone replacement therapy was not found to increase breast cancer risk when results were reported at a mean follow-up of 7.1 years.^[70] However, combination therapy with estrogen-progestin hormone replacement did demonstrate a significantly increased risk, with a hazard ratio of 1.2.^[71] This risk remained elevated for several years after discontinuation of combination therapy but then declined rapidly.^[72] After the results of the Women's Health Initiative study were made public, a subsequent decline in breast cancer incidence was noted in the US, which has been attributed, in part, to a decrease in the use of combination hormone replacement.^[73]

Hormone replacement therapy has been shown to increase breast density, which, in turn, decreases the sensitivity of mammograms.^{[74, 75]} In the Women's Health Initiative study, postmenopausal women on combination progesterone/estrogen replacement had an average increase of 6% in mammographic density, whereas their counterparts taking placebo had an approximately 1% decrease over the same period.^[76]

Environmental and occupational exposures

Geographic patterns of cancer incidence may provide insight into cancer etiology. Possible risk

factors include environmental exposures and occupational exposures from the air or water.

Workplace exposure to chemicals such as coal-tar–based products, benzene, cadmium, uranium, asbestos, or nickel can significantly increase cancer risk. For example, a significant proportion of bladder cancers may be due to chemical exposures in the aluminum, dye, paint, petroleum, rubber, and textile industries. Occupational exposures to radon and asbestos have been linked to lung cancer, and a small percentage of lung cancers are attributable to air pollution. Arsenic exposure has also been linked to increased incidence of nonmelanoma skin cancers.^[77]

The International Agency for Research on Cancer (IARC) has conducted extensive evaluations of potential carcinogens based on data from epidemiologic and animal studies. They define a carcinogenic agent as one capable of increasing the incidence of malignant neoplasms, reducing their latency, or increasing their severity or multiplicity.^[78] At this time, 108 agents are classified as group 1, or agents that are known to be carcinogenic to humans. Another 66 agents are classified as group 2A, or probable carcinogens in humans. Monographs detailing the evaluations of these agents can be accessed on the IARC website.

Infectious exposures

Approximately 17% of cancers occurring worldwide may be attributed to an infectious etiology.^[6] The primary cancers with known associations with viral infections include:

• Cervical and anogenital cancers (human papillomavirus [HPV])
• Hepatocellular carcinoma (HCC) (hepatitis B [HBV]and C [HCV])
• Kaposi sarcoma (human herpes virus [HHV]-8)
• Adult T-cell leukemia (human T-cell lymphotropic virus -1 [HTLV-I])
• Several types of non-Hodgkin lymphoma (Epstein-Barr virus and HHV-8).

Infection with human immunodeficiency virus (HIV) also increases the risk of Kaposi sarcoma and non­-Hodgkin lymphoma. In some instances, these cancers may be acquired immunodeficiency syndrome (AIDS) – defining malignancies.

Measures to prevent transmission of viral carcinogens include vaccination against carcinogenic types of HPV (see Vaccines and Medical Prevention), use of sterile needles in the healthcare and community (eg, tattoo and drug use) settings, and screening of potential blood donors. The estimated risk of hepatitis infection via blood transfusion is approximately 1 in 58,000 to 269,000 for HBV and 1 in 2 million for HCV. Risk of transmission of HTLV-1 by transfusion is 1 in 2 million, and that of HIV infection by transfusion is approximately 1 in 2 million.^{[79, 80, 81]}

For individuals who are infected with a virus that has the potential to cause cancer, several interventions are possible. Treatment of HIV infection with highly active antiretroviral therapy (HAART) can prevent lymphomas associated with this virus.^[82] Similarly, treatment of chronic HBV infection with interferon or nucleotide analogues to reduce the viral load also decreased the incidence of hepatocellular carcinoma.^[83] Alcohol avoidance may also decrease the development of hepatoma in those with chronic hepatitis.

Bacterial infection with Helicobacter pylori is associated with risk of gastric cancer. Some data suggest that eradication of H pylori infection through regimens of antibiotics and proton pump blockers may be effective as primary prevention of gastric cancer.^{[84, 85, 86]} This approach is currently being evaluated in a large-scale international clinical trial.

Sexual activity

Because some viral carcinogens are transmitted via bodily fluids, increased numbers of sexual partners or sexual contact with infected partners can increase the risk of these cancers. Although most women are exposed to HPV in their lifetime, infection is usually transient. When this virus cannot be cleared by the body, cervical cancer may develop. For this reason, individuals who have multiple sexual partners or who have a compromised immune system are more susceptible to HPV infection.

Means of reducing sexual transmission of carcinogenic viruses include awareness of the sexual and medical history of partners and use of barrier methods such as a condom during sexual intercourse.

Breast Cancer Screening

Screening for breast cancer can include different tests, of which clinical breast examination (CBE) and mammography are commonly used.

Breast self-examination

Several studies have shown that breast self-examination (BSE) does not affect breast cancer diagnosis, mortality, or stage at diagnosis.^{[87, 88]} In other studies, however, individuals who performed more thorough examinations did appear to derive a benefit.^{[89, 90]} Many experts continue to include self-examination in their recommendations due to the low cost and the benefit of making the patient an active participant in her health.

Clinical breast examination

When performed correctly by trained medical personnel in combination with mammography, clinical breast examination has greater sensitivity but a higher false-positive rate than mammography alone.^[91] One study suggested, however, that the overall effect on the cost-benefit ratio could be considered detrimental, as 55 false-positive results were identified for every cancer detected by examination that was not detected on mammography. In practice, nonstandardized breast examinations have a sensitivity of about 36% but detect about 5% of cancers not visible on mammograms.^[92] Overall, in settings in which mammography is available, clinical breast examination may provide a slight improvement in breast cancer detection; this is a cost-effective measure in countries that do not have access to imaging resources.

Mammography

Since 1990, the mortality rate from breast cancer has been declining approximately 2% per year.^[2] A study using several statistical models estimates that use of screening mammograms accounts for 28-65% of that reduction (median, 46%).^[93] A meta-analysis of randomized controlled trials showed a significant 34% reduction in breast cancer mortality at 7 years in those who had mammographic screening.^[94] A subsequent review indicated that the survival benefit from mammography was greater in women older than 50 years than in those between the ages of 40 and 49 years.^[95] However, numerous other studies have demonstrated decreased mortality in women in their 40s as well.^{[96, 97, 98, 99]} Despite this evidence of a mortality benefit, mammogram screening rates in the US decreased between 2000 and 2005.^[100]

Types of mammograms in use include film mammograms and digital mammograms. Images from digital mammograms are electronically captured and stored, whereas those from film mammograms are provided via traditional radiographic films. Overall, cancer detection rates are similar regardless of which type of mammogram is used.^{[101, 102, 103, 104, 105]} The exceptions to this rule are that digital mammography is more accurate in premenopausal women and in women with dense breasts.^[106]

Most experts agree that screening mammography should be performed routinely in women between the ages of 50 and 69 years. However, considerable controversy exists regarding the frequency of screening in this population and regarding the optimal screening approach to women outside of this age range.

For women between the ages of 50 and 69 years, some guidelines recommend annual mammography, whereas others advocate biennial or annual examination.^{[107, 108, 109, 110]} Data that directly address this question are sparse, but one observational study comparing annual and biennial screening showed no significant disadvantage from the biennial schedule in terms of detection rate or stage at time of diagnosis.^[111]

For women between the ages of 40 and 49, the breast cancer incidence and the sensitivity of mammograms are both lower than they are in women aged 50-69 years. The mammographic screening guidelines vary for this group of patients. Some guidelines, such as those from the American Cancer Society and the American College of Obstetrics and Gynecology (ACOG), advocate for annual mammograms in this population.^{[108, 109]} Other groups, such as the Canadian Task Force on the Periodic Health Examination, recommend deferring mammography until the age of 50 years in the absence of family history or other risk factors.^[112] The American College of Physicians (ACP) and the Advisory Committee on Cancer Prevention in the European Union advise discussion of screening with patients and shared decision-making.^{[113, 114]}

Data are conflicting in this area: a significant decrease in breast cancer mortality was found on meta-analysis of those who started screening in their 40s^[95] but not in a larger randomized clinical trial of annual mammography versus usual care.^[115] A cost-effectiveness study has shown that mammograms cost $21,400 per year of life saved for women aged 50-69 years, whereas the cost was $105,000 per year of life saved for women in their 40s.^[116] Although the cost for women in their 40s was 5-fold higher, both of these values fall within the accepted range for cost effectiveness.

The most recent guideline in this area was issued by the US Preventive Services Task Force (USPSTF) in 2009, which recommended against routine mammography in women aged 40-49 years, instead advocating that physicians discuss the risks and benefits of biennial screening with women in this age group.^[107] This recommendation has been met with considerable controversy, as summarized by numerous published commentaries in the literature.^{[117, 118, 119, 120, 121]}

Mammographic screening for women over the age of 70 years has also generated some controversy. Due to shorter life expectancy, the potential effect on mortality in this population is more difficult to assess. Two studies have demonstrated decreased breast cancer mortality in women aged 65 to 74 years, but these results did not achieve statistical significance in either study.^{[122, 123]} A separate study of mammographic screening in women older than age 80 years did not find an effect on rate of diagnosis, stage, or mortality.^[124] However, cost-effectiveness analysis has shown a cost per life-year gained that is within generally accepted values for continued use.^[125]

It has been suggested that bone mineral density (BMD) be incorporated into decision-making for breast screening for individuals in this 70-79 year age group, as those with higher bone density have an increased risk of breast cancer compared with those with low bone mineral density.^[126] A general rule that has been suggested for this population is that screening continue for those who have a life expectancy of 10 years or greater.

The false-positive rate of mammography is approximately 11% in the US.^[127] Risk factors for a false-positive result include younger age, previous breast biopsy, family history of breast cancer, current estrogen use, inconsistent screening, and lack of previous mammograms for comparison.^[128] Women with a false-positive result on a mammogram will be called back for additional imaging and possible biopsy of a lesion that is ultimately found to be benign.

Overdiagnosis, or screen detection of cancers that would normally not cause clinical morbidity or affect mortality, remains an issue with mammography. Estimates of overdiagnosis range from 1 in 3 to 1 in 6 cancers detected by screening mammogram.^{[129, 130]}

Some women express concern regarding the radiation associated with regular mammographic screening. Direct data regarding the risk associated with this level of radiation exposure are lacking, but a study of a risk model comparing the risk of radiation and the mortality benefit of mammographic detection concluded that the net effect of mammograms was positive for women older than age 40 years.^[131] However, it is important to note that women with BRCA mutations may be more susceptible to the effects of radiation to the breast. Studies of BRCA carriers showed that those with exposure to chest x-rays were 54% more likely to develop breast cancer; this risk increased with multiple x-rays or x-­rays done at an early age.^[132] In contrast, a study of mammograms in another BRCA population did not find an increased risk of breast cancer in a multivariate model.^[133]

Breast ultrasonography

In the US, breast ultrasonography is not usually performed as a screening study. Investigations into the combination of ultrasonography and mammography for screening have shown increased sensitivity but decreased specificity with the addition of ultrasonography.^{[134, 135, 136]} At present, although ultrasonography remains a tool for diagnosis, it is not a tool for screening of the breast.

Breast magnetic resonance imaging

At this time, breast magnetic resonance imaging (MRI) is not used for routine screening in the general population, and no studies of the effect of screening breast MRI on breast cancer mortality have been published to date. In women at high risk of breast cancer, a comparison of breast MRI and mammography showed that MRI was significantly more sensitive but less specific than mammography.^{[137, 138, 139, 140]} Women who undergo screening with breast MRI must be made aware of the increased risk of a false-positive result that requires additional studies or a biopsy.^{[138, 141]}

Guidelines published by the American Cancer Society recommend annual breast MRI (usually in addition to annual mammography) for the following patients:

• Women who are carriers of mutations in the BRCA genes or other germline mutation carriers with a known markedly increased risk of breast cancer
• First-degree relatives of mutation carriers who have not been tested themselves
• Women who have a history of radiation to the breast between the ages of 10 and 30 years
• Women with a lifetime risk of breast cancer estimated at 20% or greater according to family history–based risk assessment models (eg, BRCAPro, Myriad, Tyrer-Cusick).^[142]

The American Cancer Society guidelines state that additional data are required before making a recommendation for or against screening with breast MRI in those who have a personal history of breast cancer, those with precancerous conditions such as lobular carcinoma in situ (LCIS) or atypical hyperplasia, those with dense breasts, or those who have a 15-20% lifetime risk of breast cancer. MRls are not recommended for those with a lifetime breast cancer risk of less than 15%.

The National Institute for Health and Clinical Excellence (NICE) also issued guidelines for screening with breast MRI. These guidelines suggest annual MRI and mammography be done for the following patients^[143] :

• Women with BRCA mutations who are age 30 years and older
• Tumor-suppressor protein p53 (TP53) mutation carriers aged 20 years and older
• Women in their 30s with a 10-year risk >8%
• Women in their 40s with a 10-year risk >20%
• Women in their 40s with dense breasts and a 10-year risk >12%

Cervical Cancer Screening

The Papanicolaou (Pap) smear is the standard screening test for cervical cancer. HPV is the etiologic agent for most cervical cancers.

Pap smears

This test consists of sampling and examination of the cells at the transformation zone at the junction between the endocervix and ectocervix, which is the site of cervical dysplasia and cancers.

The efficacy of the Pap smear has never been evaluated in a randomized clinical trial, although an observational study in Canada showed a decrease in cervical and uterine cancer mortality rates in areas with high screening rates by Pap smear; areas with low rates showed an increase in mortality from these cancers.^[144]

As the use of Pap smears to screen for cervical cancer became implemented internationally, many other countries reported corresponding declines in cervical cancer incidence and mortality.^{[145, 146, 147, 148, 149, 150, 151, 152]} In addition, a case­-control study found that lack of a Pap smear within the 5 years before a cervical cancer diagnosis conveyed an almost 3-fold increase in risk of invasive cancer.^[153]

Two methods of Pap smears are in current use: the conventional smear and liquid-based cytology. The performances of these tests are similar for identifying high-grade squamous lesions with the potential to develop into cancer.^{[154, 155, 156, 157, 158, 159, 160]} Low-grade lesions and atypical squamous or glandular cells are better detected by the liquid-based technique,^{[161, 162, 163, 164, 165, 166, 167]} and a number of reviews have also found that specimen adequacy is better with the liquid-based cytology, but it is not clear that this result represents a clinically significant benefit.^{[154, 155, 161, 162, 168]} An additional benefit of the liquid-based testing is that the same specimen may be used for the Pap smear and for HPV testing (see below). Cost and availability of liquid-based testing measures are other factors that figure into the decision of which method to use.

Approximately 6-7% of Pap smears per year are read as abnormal.^{[169, 170]} Sensitivity and specificity of this test vary substantially: estimates of the sensitivity range from 30% to 87%, whereas specificity is reported as 86-100%.^[171]

Guidelines for cervical cancer screening are available from the USPSTF, the American Cancer Society, and the American College of Obstetrics and Gynecologists. All recommend starting Pap smears at age 21 years or 3 years after the onset of sexual activity. However, the guidelines vary in their recommendations for the age at which screening should be stopped: the USPSTF advises stopping at age 65 years, the American Cancer Society recommends ending screening at age 70 years if previous Pap smears have been normal, and the American College of Obstetrics and Gynecologists indicates that the age to stop screening should be determined on an individual basis.

Frequency of screening also varies among the guidelines. Data have indicated that the incidence of cervical cancer detection was only slightly lower with a screening interval of every 3 years, as compared with every year.^[172] The USPSTF advocates for screening at least every 3 years, whereas the other 2 groups suggest annual Pap smears for women younger than 30 years and screening every 2-3 years for women older than 30 years with 3 consecutive normal tests.^{[173, 174, 175]}

Women who have undergone total hysterectomy (with removal of the cervix) for benign causes do not need cervical screening; however, those who have a history of cervical or uterine cancer should continue to have Pap smears on an annual basis. Individuals with a history of immunosuppression or diethylstilbestrol (DES) exposure in utero may also require more frequent screening.

Human papillomavirus testing

Two types of tests have been developed to identify high-risk types of HPV, the etiologic agent for most cervical cancers. The first type identifies whether any of 14 high-risk HPV types are present. Examples of this type of test include the Hybrid Capture 2 and the Cervista HPV HR test. The second test, as exemplified by the Cervista HPV 16/18 test, detects HPV types 16 and 18, which are responsible for the majority of cervical cancers and high-grade cervical lesions.

HPV testing has been studied alone and in conjunction with Pap smears. This test does have a greater sensitivity than cervical cytology alone; however, its specificity is lower.^{[176, 177, 178, 179, 180, 181]}

Primary HPV testing is not recommended for women younger than 30 years, as the rate of false-positives is higher in this population. Those older than 30 years who undergo primary HPV testing require a Pap smear to follow up a positive HPV result. Consensus guidelines from the Cytopathology Education and Technology Consortium recommend that women with negative Pap smear and negative HPV testing do not require screening for at least 3 years.^[182] Those with a positive HPV test and a negative Pap smear need to repeat both tests in 1 year; if the tests are negative at that time, the women can return to screening every 3 years.

Screening recommendations by the American Cancer Society and the American College of Obstetrics and Gynecologists suggest that the screening interval may be increased to every 3 years if HPV testing is done in addition to cervical cytology. At this time, the USPSTF has made no recommendations regarding use of HPV testing.^{[173, 174, 175]}

Colon Cancer Screening

Colon cancer screening can be performed using a variety of methods; there are a total of 7 testing options recommended by the Multi-Society Task Force, which includes representation from the American Cancer Society, the American College of Radiology, and the US Multi-Society Task Force on Colorectal Cancer.^[183] The screening modalities for colon cancer can be classified into 2 general categories: stool-based tests and endoscopic or radiologic examinations.

Stool-based tests

Stool-based tests have the ability to detect early cancers and may detect advanced adenomas. The first of these tests, the guaiac-based fecal occult blood test (gFOBT), has been shown in clinical trials to reduce colon cancer mortality by up to 33% when done on an annual basis.^{[184, 185, 186, 187, 188]} Advantages of this test include its low cost and noninvasive method. However, the specificity of the gFOBT is low, giving the potential for many false-positive results that require additional testing, such as colonoscopy. Patients should avoid nonsteroidal anti-inflammatory drugs (NSAIDs), red meat, and high doses of vitamin C (>250 mg) for 48 hours before and during the testing period, and samples should be collected from 3 consecutive stools.

More recently, immunochemical-based FOBTs (also known as fecal immunochemical tests) and the stool DNA panel have also become available. The cost of the immunochemical test is greater than that of the guaiac-based test, but the specificity may be higher. Stool DNA panels are also expensive and require a greater quantity of stool to perform the test. The suggested screening interval is every 5 years.

Endoscopic and radiographic tests

Studies that image the colon and rectum have the potential to detect not only cancers but adenomatous polyps that may develop into cancers. Endoscopic tests such as flexible sigmoidoscopy and colonoscopy have been in use as screening tools for many years. Although clinical trials have not addressed the effect of these endoscopic modalities on colorectal cancer mortality in the general population, observational studies have shown that flexible sigmoidoscopy may significantly reduce deaths from colorectal cancers that occur in the area visualized by this scope.^{[189, 190, 191]}

The major limitation of flexible sigmoidoscopy is that it visualizes only the distal 60 cm of the colon. Abnormal results require examination of the entire colon by colonoscopy. The recommended screening interval for flexible sigmoidoscopy is every 5 years. Colonoscopy, however, requires significant bowel preparation and carries the risk of complications such as bleeding or bowel perforation. If there is no evidence of polyps, colonoscopy is repeated every 10 years for screening.

Double-contrast barium enema has also been included as a colon cancer screening option. Although this test is a safe means to visualize the entire bowel, its sensitivity is poor and positive results require additional evaluation with colonoscopy. The most recent USPSTF recommendations do not consider barium enema for colon cancer screening. In the absence of an abnormal result, barium enema is recommended to be performed every 5 years.

Computed tomography (CT) colonography, formerly referred to as "virtual colonoscopy," has been a more recent development. Similar to optical colonoscopy, this modality visualizes the entire bowel and requires a bowel preparation; however, CT colonography does not pose the risks of complications associated with optical colonoscopy. Positive results require colonoscopy to obtain a biopsy. There is also a risk associated with the cumulative radiation exposure for this test, which could become significant if it is performed regularly for screening. CT colonography should be performed every 5 years.

Guidelines

Several different groups have produced guidelines for colorectal cancer screening. All of the tests listed above are endorsed by the Multi-Society Task Force.^[183] This group encourages physicians and patients to discuss the screening options in terms of early detection (stool-based tests) or prevention (radiographic or endoscopic tests), and then to select a test within the chosen category. Screening is recommended to start at age 50 years in individuals who are at average risk of colon cancer and to continue until the patient's life expectancy is less than 10 years.

The recommendations of the American College of Gastroenterology differ in several respects.^[192] Although no preference for screening modality is espoused by the Multi-Society Task Force, the American College of Gastroenterology prefers colonoscopy and fecal immunochemical testing as well as recommends starting screening in black individuals at age 45 years rather than 50 years. The USPSTF guidelines prefer 3 screening strategies: FOBT every year, colonoscopy every 10 years, or the combination of flexible sigmoidoscopy every 5 years and FOBT every 3 years.^[193] The age range put forward for screening in these guidelines is age 50-75 years.

Individuals who are at increased risk of colon cancer due to a personal history of polyps, inflammatory bowel disease (IBD), or previous colorectal cancer or a family history of colorectal cancer should have screening by colonoscopy. The age at which screening is initiated may also be earlier for these individuals, based on the youngest age at which a family member was diagnosed with colorectal cancer or the time of onset of symptoms from inflammatory bowel disease. The examinations should be performed more frequently in these individuals as well, with annual colonoscopy recommended for those at genetic risk and colonoscopy every 1-2 years for those with inflammatory bowel disease. For individuals with a first-degree relative with colorectal cancer but no evidence of a genetic syndrome, screening should start at age 40 years and be repeated every 5 years.

Individuals with a personal history of adenoma should have repeat colonoscopy 3-5 years after the polyp has been removed, with the interval determined on the basis of the histology and number of polyps identified. For very large polyps that may not have been completely resected, short-term repeat colonoscopy several months later may be appropriate. Subsequent intervals for colonoscopy should be 5-10 years in individuals with polyps. For those with a personal history of colon cancer, colonoscopy should be repeated 1 year after cancer diagnosis, then in 1-3 years (based on findings at the 1-year colonoscopy). If no further pathology is identified, colonoscopy should continue at intervals of no more than 5 years for subsequent screening.^[194]

Endometrial Cancer Screening

Studies of endometrial cancer screening are lacking; however, on the basis of expert opinion, this screening is not recommended for the general population.^[195] Cervical cytology is not sensitive enough to be a reliable screening test for endometrial cancer, although it is effective in detecting pathology of the squamous cells of the cervix,^{[196, 197, 198, 199, 200]} and endometrial biopsy is an invasive procedure that often does not result in an adequate specimen. In addition, the benefit from presymptomatic diagnosis of endometrial cancer is unclear, as the majority of cancers present as dysfunctional uterine bleeding (DUB) and are diagnosed while still confined to the uterus (stage I).

Women who are at increased hereditary risk of endometrial cancer from Lynch syndrome (hereditary nonpolyposis colorectal cancer [HNPCC]) or from Cowden syndrome are recommended to undergo annual endometrial biopsy starting at age 35 years.^[195] Transvaginal ultrasonography has also been recommended for this group, although some evidence suggests that this procedure is not effective in early detection of cancers.^[201]

Lung Cancer Screening

Currently, screening for lung cancer in asymptomatic individuals is not recommended. Studies are ongoing to assess the effects of screening on lung cancer mortality. The Prostate, Lung, Colorectal, and Ovarian (PLCO) Screening Trial is investigating the use of chest x-rays as a screening tool for smokers and nonsmokers.^[202] Another study, the National Lung Cancer Screening Trial, is evaluating chest x-­rays and CT scans as screening strategies for current and former smokers.

Ovarian Cancer Screening

Detection of early stage ovarian cancer has proved difficult: unfortunately, three quarters of ovarian cancers show spread beyond the ovary at the time of diagnosis. Although current guidelines do not recommend screening for ovarian cancer for women in the general population,^{[203, 204, 205]} screening tests such as the serum biomarker CA125 and transvaginal ultrasonography are frequently used in combination to monitor women who are at increased risk of ovarian cancer due to genetic mutations in BRCA or mismatch repair genes.

Serum CA125 levels have fair sensitivity for advanced ovarian cancer but poor specificity.^[206] A small proportion of healthy women have elevated CA125 levels, and these levels can also be affected by age, smoking status, the menstrual cycle, endometriosis, cirrhosis, fibroids, pelvic inflammatory disease (PID), pleural or peritoneal fluid, or other types of cancers.^{[207, 208, 209, 210, 211, 212, 213]} A change in CA125 level over time may be a more effective means of assessing for ovarian neoplasms.

A panel of serum proteins, including CA125, was marketed in the US as a screening test for ovarian cancer under the brand name OvaSure. However, this test was designed for a population that had a much higher prevalence of ovarian cancer than that of the general population in the US.^[214] Concerns about the validity of this test in the general population led to its removal from the market.

The main imaging modality that has been evaluated for ovarian cancer screening is transvaginal ultrasonography. Although its specificity is good, the sensitivity of screening with this modality varies significantly based on the experience of the operator.^{[206, 215, 216]} Clinical trials to determine whether transvaginal ultrasonography can diagnose early stages of ovarian cancer have had conflicting results. In the National Ovarian Cancer Early Detection Program, all cancers detected in high-risk women were stage III.^[217] However, other studies conducted in lower-risk populations did have a significant proportion of early disease among the cancers detected.^{[218, 219, 220]}

The effects of screening on the early detection and mortality rates of ovarian cancer in women at average risk are being assessed in 3 large, randomized, controlled trials.^{[218, 221, 222]} Each of these studies involves multimodal screening with both transvaginal ultrasonography and serum CA125 testing, with transvaginal ultrasonography or usual care as control groups.

Because of the higher prevalence of ovarian cancer in women at high risk, the predictive value of screening is higher in this population; therefore, some expert groups do endorse screening of high-risk women. However, the screening interval is not specified by all groups. The National Comprehensive Cancer Network (NCCN) recommends that women who are at high risk of ovarian cancer due to a strong family history or a deleterious mutation should have both transvaginal ultrasonography and CA125 testing every 6 months, starting at age 35 years or 5-10 years before the earliest cancer in the family.^[223]

Prostate Cancer Screening

The mainstays of prostate cancer screening are measurement of prostate-specific antigen (PSA) and digital rectal examination (DRE). PSA testing originated as a tumor marker to assess recurrence or disease progression for men with a history of prostate cancer; however, it was adopted for screening in the general population before this use was evaluated in randomized trials. Later studies had mixed results: the European Randomized Study of Screening for Prostate Cancer (ERSPC) found a survival benefit associated with PSA screening, whereas the PLCO study did not find a benefit associated with screening by PSA and digital rectal examination.^{[224, 225]} Measurement of PSA can be affected by factors other than prostate cancer, including benign prostatic hypertrophy (BPH), ejaculation, bacterial prostatitis, urinary retention, and use of medications such as 5-alpha reductase inhibitors.^{[226, 227, 228]}

Prostate biopsy has also been shown to cause significant elevations in PSA sustained for up to 1 month, whereas digital rectal examination is not felt to have a substantive effect on PSA measurements.^{[229, 230]} Overall, when abnormal PSA values are defined as those over 4.0 ng/mL, the sensitivity of the PSA test is 70-80% and the specificity is 60-70%.^[231] A noteworthy exception is the Prostate Cancer Prevention Trial, which showed a sensitivity of 21% for PSA at this level.^[232]

Several approaches have been taken to try to improve the utility of PSA measurement, including use of PSA velocity to assess change over time, PSA density to adjust values according to prostate volume as measured by ultrasonography or MRI, the ratio of free to total PSA, and complexed PSA, as well as individualized reference ranges based on age or race. However, these modified values have not replaced standard PSA testing in most practices.

Despite the long-standing use of digital examination of the prostate through the rectum (ie, digital rectal examination), there have been no controlled studies that demonstrated an effect of this test on prostate cancer mortality or incidence.^[233] Assessment of the prostate by this technique often results in inconsistent results, even among specialists.^[234] The sensitivity has been estimated at 59% and the specificity at 94%.^[235] Clinically, the overall sensitivity is improved when digital rectal examination and PSA are used in combination, although PSA appears to outperform digital rectal examination in some aspects.^{[195, 236, 237]}

Guidelines regarding screening for prostate cancer with PSA testing and digital rectal examination vary considerably. Several groups, including the American Cancer Society, the American College of Physicians, and the American Urological Association, recommend discussion of the risks and benefits of screening with the patient and individualizing decisions.^{[195, 238, 239]} The American Cancer Society and American Urological Association recommend consideration of starting screening at age 50 years for most men and at age 45 years for men who are at increased risk of prostate cancer due to family history or ethnic background. Their regimen consists of annual PSA testing and digital rectal examination for all men who have a life expectancy of at least 10 years and who desire screening.

Numerous other groups, including the USPSTF, the Canadian Task Force on Preventive Health Care, and the European Union, find the evidence insufficient to support screening at this time.^{[114, 240, 241]} The USPSTF feels that the potential harms associated with screening individuals older than 75 years exceed the potential benefits, and the Canadian Task Force recommends against use of PSA for screening in any age group.

Breast Cancer

At present, 2 medications are available for the primary prevention of breast cancer: tamoxifen and raloxifene. These drugs are selective estrogen receptor modulators (SERMs) that have both estrogen agonist and antagonist effects.

Tamoxifen

The Breast Cancer Prevention Trial (BCPT) found that, compared with patients receiving placebo, tamoxifen users had a 50% reduction in the incidence of invasive and noninvasive estrogen receptor–positive breast cancers. The investigators evaluated the use of tamoxifen for the prevention of breast cancer in women at increased risk based on age older than 60 years, personal history of lobular carcinoma in situ, or a 5-year Gail score of more than 1.66%.^[242] On long-­term follow-up, the protective effect of tamoxifen persisted, with approximately 43% reduction in invasive cancer risk.^[243] However, tamoxifen was also shown to produce an approximately 2-fold increased risk of endometrial cancer and thromboembolic events, including stroke.

A second study, the International Breast Cancer Intervention Study (IBIS-l), also found a significant reduction in the incidence of breast cancer in tamoxifen users, but the all-cause mortality rate was higher in the tamoxifen intervention group.^[244] A meta-analysis concluded that use of tamoxifen reduced breast cancer risk by 38% and doubled the risk of endometrial cancer and thromboembolic events, but it did not affect overall mortality.^[245]

The USPSTF does not recommend use of tamoxifen for chemoprevention of breast cancer in women of average risk. For those who have an increased risk due to family history, BRCA or other mutation carrier status, or conditions such as lobular carcinoma in situ, the benefits of treatment must be balanced against the risks of side effects on an individual basis.^[246]

Raloxifene

In the study of tamoxifen and raloxifene (STAR) trial, in which raloxifene (Evista) was compared with tamoxifen for prevention of breast cancer in postmenopausal women, the incidence rates for invasive breast cancer were similar with both drugs, but the risks of thromboembolic adverse effects and endometrial cancer were significantly lower with raloxifene.^[247] In addition, although the rate of noninvasive breast cancer was higher in the raloxifene group than in the tamoxifen group, this difference did not achieve statistical significance. The profile of other side effects also differed between the 2 groups: tamoxifen was associated with higher rates of gynecologic problems, vasomotor symptoms, and bladder symptoms, whereas sexual dysfunction, musculoskeletal issues, and weight gain were more frequent with raloxifene.^[248]

The US Food and Drug Administration (FDA) approved raloxifene for use for breast cancer chemoprevention in postmenopausal women. It is important to note that this drug has not been studied in premenopausal women and is not approved for use in this population. To date, there are no reports on the use of raloxifene in BRCA mutation carriers.

Cervical Cancer

Two vaccines have been developed against HPV, the etiologic agent of most cases of cervical cancer: Gardasil (Merck & Co, Inc. Whitehouse Station, NJ) and Cervarix (GlaxoSmithKline Biologicals, Rixensart, Belgium).

Gardasil and Cervarix

Gardasil is a quadrivalent vaccine that acts against HPV genotypes 6, 11, 16, and 18. The HPV-16 and HPV-18 genotypes cause almost two thirds of all cervical cancers and cervical intraepithelial neoplasia (CIN) 2 and 3, whereas the the HPV-6 and HPV-11 genotypes are implicated in genital warts. Cervarix, is a bivalent vaccine that also targets HPV-16 and HPV-18.

Both vaccines are given in 3 doses over a 6-month period. Each of the vaccines has been shown to be very effective in the prevention of CIN 2/3 or cervical cancers in subjects who had not been previously exposed to HPV. However, neither vaccine showed strong efficacy in preventing premalignant or malignant cervical disease in subjects who were not HPV-naive; nor did the vaccines show efficacy in the treatment of existing HPV infections.^{[249, 250]}

Because cervical cancers can be caused by HPV genotypes other than those contained in the vaccines, cervical screening with Pap smears should be continued after vaccination. However, the duration of protection against HPV after vaccination remains unknown.

Side effects

The most frequent side effect associated with vaccination was a mild reaction at the injection site.^{[249, 250]} Postlicensure surveillance reports for Gardasil also indicated increased risk of post­vaccination syncope and thromboembolic events; however, nearly all thromboembolic events occurred in those with a known risk factor, such as oral contraceptive use or family history of thromboembolic disease.^[251] At this time, postlicensure data are not available for Cervarix.

Recommendations

Several groups have made recommendations regarding the appropriate use of HPV vaccinations. The Advisory Committee on Immunization Practices (ACIP) and the American College of Obstetricians and Gynecologists recommend that vaccination should be offered to girls before their becoming sexually active (ACIP: age 11-12 y; ACOG: age 13-15 y).^{[252, 253]} Both groups advocate "catch-up" vaccination for females between the ages of 13 and 26 years if they have not had previous vaccination, and they indicate that vaccination may occur as early as age 9 years. HPV testing is not recommended before vaccination.

The WHO also recommends vaccinations for girls between the ages of 9 and 13 years^[254] ; however, the American Cancer Society recommends girls between the ages of 11 and 18 years be offered vaccination.^[255] These guidelines did not support catch-up vaccination, indicating that additional evidence of benefit was needed for those older than 18 years.

Male vaccination for HPV

At present, HPV vaccination is not recommended for boys or men. However, vaccination of males remains a consideration, both to prevent reinfection of their female sexual partners and to prevent genital warts and anal or penile cancers. Data for this indication are under review at the FDA.

Immunosuppressed/immunocompromised individuals

HPV vaccination has not been studied in immunosuppressed or immunocompromised individuals; therefore, at this time, vaccination is not specifically recommended for those in this group.

Prostate Cancer

5-alpha reductase inhibitors have been shown to have protective effects against prostate cancer in clinical trials. Meta-analysis of 5-alpha reductase inhibitor studies indicates that these agents decrease the risk of prostate cancer by approximately 25%.^[256]

Finasteride

The Prostate Cancer Prevention Trial (PCPT), which compared finasteride to placebo in nearly 19,000 men who were at increased risk of prostate cancer due to age, ethnic background, or family history of prostate cancer, was closed early due to a 25% decrease in numbers of prostate cancer in the finasteride group relative to the control group. However, the finasteride group was noted to have more aggressive cancers (Gleason score >6).^[257]

Several possible explanations for this result have been presented, including diagnostic bias due to decreased prostate volume in the participants taking finasteride and detection bias due to the effects of finasteride on PSA testing.^{[258, 259, 260, 261, 262, 263]} Finasteride decreases PSA levels by 50% on average.^{[264, 265, 266]}

Side effects included gynecomastia and sexual dysfunction, which was mild and decreased over time. It remains unclear whether finasteride decreases mortality from prostate cancer.

Dutasteride

Preliminary results from the REDUCE (REduction by DUtasteride of prostate Cancer Events) trial, which investigated dutasteride, another 5-alpha reductase inhibitor, as a chemopreventive agent for men at increased risk of prostate cancer showed a 23% decrease in prostate cancer incidence in the dutasteride arm, which achieved statistical significance.^[267] Unlike the PCPT trial, the incidence of high-grade cancers was not increased in the REDUCE study. Final results on this study are pending.

Recommendations

Guidelines from the American Society of Clinical Oncology and the American Urologic Association recommend that men discuss the risks and benefits of chemoprevention with these medications with their physicians.^[268]

For individuals at very high risk of cancer, such as those with hereditary genetic predisposition, surgical intervention provides another means of risk reduction. Main examples of surgical management of cancer risk include prophylactic mastectomy in BRCA mutation carriers, prophylactic salpingo-­oophorectomy in BRCA carriers and mismatch repair gene mutation carriers, and prophylactic colectomy in individuals with familial adenomatous polyposis (FAP).

Prophylactic mastectomy

Prophylactic mastectomy has been a mainstay of management of BRCA mutation carriers because of their markedly increased risk of breast cancer. The breast cancer risk reduction associated with bilateral prophylactic mastectomy is approximately 90%.^{[269, 270, 271, 272, 273]} Simple or skin-sparing mastectomy techniques are preferred over subcutaneous mastectomy, because the latter technique tends to leave more residual breast tissue with the potential to develop cancer.^{[274, 275, 276]}

Despite the considerable risk reduction associated with this procedure, utilization of prophylactic mastectomy remains far less than that of prophylactic salpingo-oophorectomy. Possible reasons for the discrepancy between the rates of these prophylactic surgeries include lack of data proving survival benefit, concerns about appearance and sexuality following mastectomy, availability of medications that reduce breast cancer risk, and the options of screening modalities that can detect breast cancer at a premalignant or early stage.

Prophylactic salpingo-oophorectomy

Prophylactic salpingo-oophorectomy has primarily been studied in women with BRCA1 or BRCA2 germline mutations, although it is also appropriate for women with Lynch syndrome (hereditary non­polyposis colorectal cancer) and some other hereditary syndromes. This procedure provides premenopausal BRCA mutation carriers with protection from both ovarian cancer (90-95% risk reduction) and breast cancer (50% risk reduction); for postmenopausal carriers, it provides protection from ovarian cancer only.^{[277, 278, 279, 280, 281]} Even after oophorectomy, however, a small risk of primary peritoneal cancer remains.^{[282, 283]}

Prophylactic gynecologic surgery has been shown to improve overall and cancer-related survival in this population.^[284] Similarly, risk-reducing surgery dramatically decreases the incidence of cancer in the Lynch syndrome population as well.^[285]

For BRCA mutation carriers, experts recommend that bilateral salingo-oophorectomy take place between the ages of 35 and 40 years or after childbearing is complete. The fallopian tubes should be removed as well, due to the increased incidence of fallopian tube carcinoma in this population. Careful pathology review of both the ovaries and fallopian tubes is vital to detect clinically occult cancers, which is reported in approximately 4-8% of women undergoing this procedure.^{[286, 287, 288]}

Prophylactic hysterectomy at the time of oophorectomy remains a point of some debate. For women with Lynch syndrome, the risk of endometrial cancer is also increased, and it is appropriate to remove the uterus at the time of oophorectomy. Although BRCA mutations are not generally felt to increase the risk of endometrial cancer, some data suggest that the incidence of these cancers is higher than would be expected in this population.^{[289, 290]} These results may be due to use of tamoxifen for chemoprevention of breast cancer. Removal of the uterus could allow for more complete removal of the fallopian tubes and as well as use of unopposed estrogen for hormone replacement therapy.

Another issue of debate is the use of hormone replacement therapy after prophylactic oophorectomy, which remains controversial in this population due to the risks of breast cancer. At least one report has found that short-term use of hormone replacement to treat menopausal symptoms after oophorectomy in BRCA carriers did not increase the incidence of cancers.^[291] However, long-term follow-up data are lacking, so careful counseling regarding the risks and benefits of hormone replacement therapy is appropriate.

Colectomy

Individuals with the hereditary syndrome familial adenomatous polyposis are often afflicted with hundreds to thousands of colorectal polyps and have a virtual certainty of developing colorectal cancer in their lifetime if their disease is unchecked. A standard risk-reducing measure in this population is prophylactic colectomy, which is generally undertaken at the appearance of adenomas in known mutation carriers. Depending on the phenotype of the individual and other affected members of the family, surgical options include total proctocolectomy with ileal pouch and anastomosis, total abdominal colectomy with ileorectal anastomosis, or total proctocolectomy with ileostomy. For individuals who have an intact rectum after surgery, regular lower endoscopic surveillance is still recommended; similarly, surveillance of the ileal pouch should occur every 2 years in individuals who have undergone this surgical option.^{[292, 293]}

Although the burden of cancer is daunting, primary and secondary prevention of this disease is possible in many instances. Recent trends for decreased cancer incidence and mortality in the US are largely due to improvements in risk reduction measures and early detection of cancer through screening.