Cleveland Clinic Journal of Medicine

Sexually transmitted infections (STIs) such as gonorrhea, chlamydia, and syphilis are still increasing in incidence and probably will continue to do so in the near future. Moreover, drug-resistant strains of Neisseria gonorrhoeae are emerging, as are less-known organisms such as Mycoplasma genitalium.

Now the good news: new tests for STIs are available or are coming! Based on nucleic acid amplification, these tests can be performed at the point of care, so that patients can leave the clinic with an accurate diagnosis and proper treatment for themselves and their sexual partners. Also, the tests can be run on samples collected by the patients themselves, either swabs or urine collections, eliminating the need for invasive sampling and making doctor-shy patients more likely to come in to be treated.¹ We hope that by using these sensitive and accurate tests we can begin to bend the upward curve of STIs and be better antimicrobial stewards.²

This article reviews current issues surrounding STI control, and provides detailed guidance on recognizing, testing for, and treating gonorrhea, chlamydia, trichomoniasis, and M genitalium infection.

STI RATES ARE HIGH AND RISING

STIs are among the most common acute infectious diseases worldwide, with an estimated 1 million new curable cases every day.³ Further, STIs have major impacts on sexual, reproductive, and psychological health.

In the United States, rates of reportable STIs (chlamydia, gonorrhea, and syphilis) are rising.⁴ In addition, more-sensitive tests for trichomoniasis, which is not a reportable infection in any state, have revealed it to be more prevalent than previously thought.⁵

BARRIERS AND CHALLENGES TO DIAGNOSIS

The medical system does not fully meet the needs of some populations, including young people and men who have sex with men, regarding their sexual and reproductive health. 

Ongoing barriers among young people include reluctance to use available health services, limited access to STI testing, worries about confidentiality, and the shame and stigma associated with STIs.⁶

Men who have sex with men have a higher incidence of STIs than other groups. Since STIs are associated with a higher risk of human immunodeficiency virus (HIV) infection, it is important to detect, diagnose, and manage STIs in this group—and in all high-risk groups. Rectal STIs are an independent risk factor for incident HIV infection.⁷ In addition, many men who have sex with men face challenges navigating the emotional, physical, and cognitive aspects of adolescence, a voyage further complicated by mental health issues, unprotected sexual encounters, and substance abuse in many, especially among minority youth.⁸ These same factors also impair their ability to access resources for preventing and treating HIV and other STIs.

STI diagnosis is often missed

Most people who have STIs feel no symptoms, which increases the importance of risk-based screening to detect these infections.^9,10 In many other cases, STIs manifest with nonspecific genitourinary symptoms that are mistaken for urinary tract infection. Tomas et al¹¹ found that of 264 women who presented to an emergency department with genitourinary symptoms or were being treated for urinary tract infection, 175 were given a diagnosis of a urinary tract infection. Of these, 100 (57%) were treated without performing a urine culture; 60 (23%) of the 264 women had 1 or more positive STI tests, 22 (37%) of whom did not receive treatment for an STI.

Poor follow-up of patients and partners

Patients with STIs need to be retested 3 months after treatment to make sure the treatment was effective. Another reason for follow-up is that these patients are at higher risk of another infection within a year.¹²

Although treating patients’ partners has been shown to reduce reinfection rates, fewer than one-third of STIs (including HIV infections) were recognized through partner notification between 2010 and 2012 in a Dutch study, in men who have sex with men and in women.¹³ Challenges included partners who could not be identified among men who have sex with men, failure of heterosexual men to notify their partners, and lower rates of partner notification for HIV.  

In the United States, “expedited partner therapy” allows healthcare providers to provide a prescription or medications to partners of patients diagnosed with chlamydia or gonorrhea without examining the partner.¹⁴ While this approach is legal in most states, implementation can be challenging.¹⁵

STI EVALUATION

History and physical examination

A complete sexual history helps in estimating the patient’s risk of an STI and applying appropriate risk-based screening. Factors such as sexual practices, use of barrier protection, and history of STIs should be discussed.

Physical examination is also important. Although some patients may experience discomfort during a genital or pelvic examination, omitting this step may lead to missed diagnoses in women with STIs.¹⁶

Laboratory testing

Laboratory testing for STIs helps ensure accurate diagnosis and treatment. Empiric treatment without testing could give a patient a false sense of health by missing an infection that is not currently causing symptoms but that could later worsen or have lasting complications. Failure to test patients also misses the opportunity for partner notification, linkage to services, and follow-up testing.

Many of the most common STIs, including gonorrhea, chlamydia, and trichomoniasis, can be detected using vaginal, cervical, or urethral swabs or first-catch urine (from the initial urine stream). In studies that compared various sampling methods,¹⁷ self-collected urine samples for gonorrhea in men were nearly as good as clinician-collected swabs of the urethra. In women, self-collected vaginal swabs for gonorrhea and chlamydia were nearly as good as clinician-collected vaginal swabs. While urine specimens are acceptable for chlamydia testing in women, their sensitivity may be slightly lower than with vaginal and endocervical swab specimens.^18,19

A major advantage of urine specimens for STI testing is that collection is noninvasive and is therefore more likely to be acceptable to patients. Urine testing can also be conducted in a variety of nonclinical settings such as health fairs, pharmacy-based screening programs, and express STI testing sites, thus increasing availability.

To prevent further transmission and morbidity and to aid in public health efforts, it is critical to recognize the cause of infectious cervicitis and urethritis and to screen for STIs according to guidelines.¹² Table 1 summarizes current screening and laboratory testing recommendations.

Gonorrhea and chlamydia are the 2 most frequently reported STIs in the United States, with more than 550,000 cases of gonorrhea and 1.7 million cases of chlamydia reported in 2017.⁴

Both infections present similarly: cervicitis or urethritis characterized by discharge (mucopurulent discharge with gonorrhea) and dysuria. Untreated, they can lead to pelvic inflammatory disease, inflammation, and infertility.

Extragenital infections can be asymptomatic or cause exudative pharyngitis or proctitis. Most people in whom chlamydia is detected from pharyngeal specimens are asymptomatic. When pharyngeal symptoms exist secondary to gonorrheal infection, they typically include sore throat and pharyngeal exudates. However, Komaroff et al,²⁰ in a study of 192 men and women who presented with sore throat, found that only 2 (1%) tested positive for N gonorrhoeae.

Screening for gonorrhea and chlamydia

Best practices include screening for gonorrhea and chlamydia as follows^21–23:

• Every year in sexually active women through age 25 (including during pregnancy) and in older women who have risk factors for infection¹²
• At least every year in men who have sex with men, at all sites of sexual contact (urethra, pharynx, rectum), along with testing for HIV and syphilis
• Every 3 to 6 months in men who have sex with men who have multiple or anonymous partners, who are sexually active and use illicit drugs, or who have partners who use illicit drugs
• Possibly every year in young men who live in high-prevalence areas or who are seen in certain clinical settings, such as STI and adolescent clinics.

Specimens. A vaginal swab is preferred for screening in women. Several studies have shown that self-collected swabs have clinical sensitivity and specificity comparable to that of provider-collected samples.^17,24 First-catch urine or endocervical swabs have similar performance characteristics and are also acceptable. In men, urethral swabs or first-catch urine samples are appropriate for screening for urogenital infections.

Testing methods. Testing for both pathogens should be done simultaneously with a nucleic acid amplification test (NAAT). Commercially available NAATs are more sensitive than culture and antigen testing for detecting gonorrhea and chlamydia.^25–27

Most assays are approved by the US Food and Drug Administration (FDA) for testing vaginal, urethral, cervical, and urine specimens. Until recently, no commercial assay was cleared for testing extragenital sites, but recommendations for screening extragenital sites prompted many clinical laboratories to validate throat and rectal swabs for use with NAATs, which are more sensitive than culture at these sites.^25,28 The recent FDA approval of extragenital specimen types for 2 commercially available assays may increase the availability of testing for these sites.

Data on the utility of NAATs for detecting chlamydia and gonorrhea in children are limited, and many clinical laboratories have not validated molecular methods for testing in children. Current guidelines specific to this population should be followed regarding test methods and preferred specimen types.^12,29,30

Although gonococcal infection is usually diagnosed with culture-independent molecular methods, antimicrobial resistance is emerging. Thus, failure of the combination of ceftriaxone and azithromycin should prompt culture-based follow-up testing to determine antimicrobial susceptibility.

Strategies for treatment and control

Historically, people treated for gonorrhea have been treated for chlamydia at the same time, as these diseases tend to go together. This can be with a single intramuscular dose of ceftriaxone for the gonorrhea plus a single oral dose of azithromycin for the chlamydia.¹² For patients who have only gonorrhea, this double regimen may help prevent the development of resistant gonorrhea strains.

All the patient’s sexual partners in the previous 60 days should be tested and treated, and expedited partner therapy should be offered if possible. Patients should be advised to have no sexual contact until they complete the treatment, or 7 days after single-dose treatment. Testing should be repeated 3 months after treatment.

A member of the Mycoplasmataceae family, M genitalium was originally identified as a pathogen in the early 1980s but has only recently emerged as an important cause of STI. Studies indicate that it is responsible for 10% to 20% of cases of nongonococcal urethritis and 10% to 30% of cases of cervicitis.^31–33 Additionally, 2% to 22% of cases of pelvic inflammatory disease have evidence of M genitalium.^34,35

However, data on M genitalium prevalence are suspect because the organism is hard to identify—lacking a cell wall, it is undetectable by Gram stain.³⁶ Although it has been isolated in respiratory and synovial fluids, it has so far been recognized to be clinically important only in the urogenital tract. It can persist for years in infected patients by exploiting specialized cell-surface structures to invade cells.³⁶ Once inside a cell, it triggers secretion of mycoplasmal toxins and destructive metabolites such as hydrogen peroxide, evading the host immune system as it does so.³⁷

Testing guidelines for M genitalium

Current guidelines do not recommend routine screening for M genitalium, and no commercial test was available until recently.¹² Although evidence suggests that M genitalium is independently associated with preterm birth and miscarriages,³⁸ routine screening of pregnant women is not recommended.¹²

Testing for M genitalium should be considered in cases of persistent or recurrent nongonococcal urethritis in patients who test negative for gonorrhea and chlamydia or for whom treatment has failed.¹² Many isolates exhibit genotypic resistance to macrolide antibiotics, which are often the first-line therapy for nongonococcal urethritis.³⁹

Further study is needed to evaluate the potential impact of routine screening for M genitalium on the reproductive and sexual health of at-risk populations.

Diagnostic tests for M genitalium

Awareness of M genitalium as a cause of nongonococcal urethritis has been hampered by a dearth of diagnostic tests.⁴⁰ The organism’s fastidious requirements and extremely slow growth preclude culture as a practical method of diagnosis.⁴¹ Serologic assays are dogged by cross-reactivity and poor sensitivity.^42,43 Thus, molecular assays for detecting M genitalium and associated resistance markers are preferred for diagnosis.¹²

Several molecular tests are approved, available, and in use in Europe for diagnosing M genitalium infection,⁴⁰ and in January 2019 the FDA approved a molecular test that can detect M genitalium in urine specimens and vaginal, endocervical, urethral, and penile meatal swabs. Although vaginal swabs are preferred for this assay because they have higher sensitivity (92% for provider-collected and 99% for patient-collected swabs), urine specimens are acceptable, with a sensitivity of 78%.⁴⁴

At least 1 company is seeking FDA clearance for another molecular diagnostic assay for detecting M genitalium and markers of macrolide resistance in urine and genital swab specimens. Such assays may facilitate appropriate treatment.

Clinicians should stay abreast of diagnostic testing options, which are likely to become more readily available soon.

A high rate of macrolide resistance

Because M genitalium lacks a cell wall, antibiotics such as beta-lactams that target cell wall synthesis are ineffective.

Regimens for treating M genitalium are outlined in Table 2.¹² Azithromycin is more effective than doxycycline. However, as many as 50% of strains were macrolide-resistant in a cohort of US female patients.⁴⁵ Given the high incidence of treatment failure with azithromycin 1 g, it is thought that this regimen might select for resistance. For cases in which symptoms persist, a 1- to 2-week course of moxifloxacin is recommended.¹² However, this has not been validated by clinical trials, and failures of the 7-day regimen have been reported.⁴⁶

Partners of patients who test positive for M genitalium should also be tested and undergo clinically applicable screening for nongonococcal urethritis, cervicitis, and pelvic inflammatory disease.¹²

TRICHOMONIASIS

Trichomoniasis, caused by the parasite Trichomonas vaginalis, is the most prevalent nonviral STI in the United States. It disproportionately affects black women, in whom the prevalence is 13%, compared with 1% in non-Hispanic white women.⁴⁷ It is also present in 26% of women with symptoms who are seen in STI clinics and is highly prevalent in incarcerated populations. It is uncommon in men who have sex with men.⁴⁸

In men, trichomoniasis manifests as urethritis, epididymitis, or prostatitis. While most infected women have no symptoms, they may experience vaginitis with discharge that is diffuse, frothy, pruritic, malodorous, or yellow-green. Vaginal and cervical erythema (“strawberry cervix”) can also occur.

Screening for trichomoniasis

Current guidelines of the US Centers for Disease Control and Prevention (CDC) recommend testing for T vaginalis in women who have symptoms and routinely screening in women who are HIV-positive, regardless of symptoms. There is no evidence to support routine screening of pregnant women without symptoms, and pregnant women who do have symptoms should be evaluated according to the same guidelines as for nonpregnant women.¹² Testing can be considered in patients who have no symptoms but who engage in high-risk behaviors and in areas of high prevalence.

A lack of studies using sensitive methods for T vaginalis detection has hampered a true estimation of disease burden and at-risk populations. Screening recommendations may evolve in upcoming clinical guidelines as the field advances.

As infection can recur, women should be retested 3 months after initial diagnosis.¹²

NAAT is the preferred test for trichomoniasis

Commercially available diagnostic tests for trichomoniasis include culture, antigen testing, and NAAT.⁴⁹ While many clinicians do their own wet-mount microscopy for a rapid result, this method has low sensitivity.⁵⁰ Similarly, antigen testing and culture perform poorly compared with NAATs, which are the gold standard for detection.^51,52 A major advantage of NAATs for T vaginalis detection is that they combine high sensitivity and fast results, facilitating diagnosis and appropriate treatment of patients and their partners.

In spite of these benefits, adoption of molecular diagnostic testing for T vaginalis has lagged behind that for chlamydia and gonorrhea.⁵³ FDA-cleared NAATs are available for testing vaginal, cervical, or urine specimens from women, but until recently, there were no approved assays for testing in men. The Cepheid Xpert TV assay, which is valid for male urine specimens to diagnose other sexually transmitted diseases, has demonstrated excellent diagnostic sensitivity for T vaginalis in men and women.⁵⁴ Interestingly, a large proportion of male patients in this study had no symptoms, suggesting that screening of men in high-risk groups may be warranted.

7-day metronidazole treatment beats single-dose treatment

The first-line treatment for trichomoniasis has been a single dose of metronidazole 2 g by mouth, but in a recent randomized controlled trial,⁵⁵ a course of 500 mg by mouth twice a day for 7 days was 45% more effective at 4 weeks than a single dose, and it should now be the preferred regimen.

In clinical trials,⁵⁶ a single dose of tinidazole 2 g orally was equivalent or superior to metronidazole 2 g and had fewer gastrointestinal side effects, but it is more expensive.

Sexually transmitted infections (STIs) such as gonorrhea, chlamydia, and syphilis are still increasing in incidence and probably will continue to do so in the near future. Moreover, drug-resistant strains of Neisseria gonorrhoeae are emerging, as are less-known organisms such as Mycoplasma genitalium.

Now the good news: new tests for STIs are available or are coming! Based on nucleic acid amplification, these tests can be performed at the point of care, so that patients can leave the clinic with an accurate diagnosis and proper treatment for themselves and their sexual partners. Also, the tests can be run on samples collected by the patients themselves, either swabs or urine collections, eliminating the need for invasive sampling and making doctor-shy patients more likely to come in to be treated.¹ We hope that by using these sensitive and accurate tests we can begin to bend the upward curve of STIs and be better antimicrobial stewards.²

This article reviews current issues surrounding STI control, and provides detailed guidance on recognizing, testing for, and treating gonorrhea, chlamydia, trichomoniasis, and M genitalium infection.

STI RATES ARE HIGH AND RISING

STIs are among the most common acute infectious diseases worldwide, with an estimated 1 million new curable cases every day.³ Further, STIs have major impacts on sexual, reproductive, and psychological health.

In the United States, rates of reportable STIs (chlamydia, gonorrhea, and syphilis) are rising.⁴ In addition, more-sensitive tests for trichomoniasis, which is not a reportable infection in any state, have revealed it to be more prevalent than previously thought.⁵

BARRIERS AND CHALLENGES TO DIAGNOSIS

The medical system does not fully meet the needs of some populations, including young people and men who have sex with men, regarding their sexual and reproductive health. 

Ongoing barriers among young people include reluctance to use available health services, limited access to STI testing, worries about confidentiality, and the shame and stigma associated with STIs.⁶

Men who have sex with men have a higher incidence of STIs than other groups. Since STIs are associated with a higher risk of human immunodeficiency virus (HIV) infection, it is important to detect, diagnose, and manage STIs in this group—and in all high-risk groups. Rectal STIs are an independent risk factor for incident HIV infection.⁷ In addition, many men who have sex with men face challenges navigating the emotional, physical, and cognitive aspects of adolescence, a voyage further complicated by mental health issues, unprotected sexual encounters, and substance abuse in many, especially among minority youth.⁸ These same factors also impair their ability to access resources for preventing and treating HIV and other STIs.

STI diagnosis is often missed

Most people who have STIs feel no symptoms, which increases the importance of risk-based screening to detect these infections.^9,10 In many other cases, STIs manifest with nonspecific genitourinary symptoms that are mistaken for urinary tract infection. Tomas et al¹¹ found that of 264 women who presented to an emergency department with genitourinary symptoms or were being treated for urinary tract infection, 175 were given a diagnosis of a urinary tract infection. Of these, 100 (57%) were treated without performing a urine culture; 60 (23%) of the 264 women had 1 or more positive STI tests, 22 (37%) of whom did not receive treatment for an STI.

Poor follow-up of patients and partners

Patients with STIs need to be retested 3 months after treatment to make sure the treatment was effective. Another reason for follow-up is that these patients are at higher risk of another infection within a year.¹²

Although treating patients’ partners has been shown to reduce reinfection rates, fewer than one-third of STIs (including HIV infections) were recognized through partner notification between 2010 and 2012 in a Dutch study, in men who have sex with men and in women.¹³ Challenges included partners who could not be identified among men who have sex with men, failure of heterosexual men to notify their partners, and lower rates of partner notification for HIV.  

In the United States, “expedited partner therapy” allows healthcare providers to provide a prescription or medications to partners of patients diagnosed with chlamydia or gonorrhea without examining the partner.¹⁴ While this approach is legal in most states, implementation can be challenging.¹⁵

STI EVALUATION

History and physical examination

A complete sexual history helps in estimating the patient’s risk of an STI and applying appropriate risk-based screening. Factors such as sexual practices, use of barrier protection, and history of STIs should be discussed.

Physical examination is also important. Although some patients may experience discomfort during a genital or pelvic examination, omitting this step may lead to missed diagnoses in women with STIs.¹⁶

Laboratory testing

Laboratory testing for STIs helps ensure accurate diagnosis and treatment. Empiric treatment without testing could give a patient a false sense of health by missing an infection that is not currently causing symptoms but that could later worsen or have lasting complications. Failure to test patients also misses the opportunity for partner notification, linkage to services, and follow-up testing.

Many of the most common STIs, including gonorrhea, chlamydia, and trichomoniasis, can be detected using vaginal, cervical, or urethral swabs or first-catch urine (from the initial urine stream). In studies that compared various sampling methods,¹⁷ self-collected urine samples for gonorrhea in men were nearly as good as clinician-collected swabs of the urethra. In women, self-collected vaginal swabs for gonorrhea and chlamydia were nearly as good as clinician-collected vaginal swabs. While urine specimens are acceptable for chlamydia testing in women, their sensitivity may be slightly lower than with vaginal and endocervical swab specimens.^18,19

A major advantage of urine specimens for STI testing is that collection is noninvasive and is therefore more likely to be acceptable to patients. Urine testing can also be conducted in a variety of nonclinical settings such as health fairs, pharmacy-based screening programs, and express STI testing sites, thus increasing availability.

To prevent further transmission and morbidity and to aid in public health efforts, it is critical to recognize the cause of infectious cervicitis and urethritis and to screen for STIs according to guidelines.¹² Table 1 summarizes current screening and laboratory testing recommendations.

Gonorrhea and chlamydia are the 2 most frequently reported STIs in the United States, with more than 550,000 cases of gonorrhea and 1.7 million cases of chlamydia reported in 2017.⁴

Both infections present similarly: cervicitis or urethritis characterized by discharge (mucopurulent discharge with gonorrhea) and dysuria. Untreated, they can lead to pelvic inflammatory disease, inflammation, and infertility.

Extragenital infections can be asymptomatic or cause exudative pharyngitis or proctitis. Most people in whom chlamydia is detected from pharyngeal specimens are asymptomatic. When pharyngeal symptoms exist secondary to gonorrheal infection, they typically include sore throat and pharyngeal exudates. However, Komaroff et al,²⁰ in a study of 192 men and women who presented with sore throat, found that only 2 (1%) tested positive for N gonorrhoeae.

Screening for gonorrhea and chlamydia

Best practices include screening for gonorrhea and chlamydia as follows^21–23:

• Every year in sexually active women through age 25 (including during pregnancy) and in older women who have risk factors for infection¹²
• At least every year in men who have sex with men, at all sites of sexual contact (urethra, pharynx, rectum), along with testing for HIV and syphilis
• Every 3 to 6 months in men who have sex with men who have multiple or anonymous partners, who are sexually active and use illicit drugs, or who have partners who use illicit drugs
• Possibly every year in young men who live in high-prevalence areas or who are seen in certain clinical settings, such as STI and adolescent clinics.

Specimens. A vaginal swab is preferred for screening in women. Several studies have shown that self-collected swabs have clinical sensitivity and specificity comparable to that of provider-collected samples.^17,24 First-catch urine or endocervical swabs have similar performance characteristics and are also acceptable. In men, urethral swabs or first-catch urine samples are appropriate for screening for urogenital infections.

Testing methods. Testing for both pathogens should be done simultaneously with a nucleic acid amplification test (NAAT). Commercially available NAATs are more sensitive than culture and antigen testing for detecting gonorrhea and chlamydia.^25–27

Most assays are approved by the US Food and Drug Administration (FDA) for testing vaginal, urethral, cervical, and urine specimens. Until recently, no commercial assay was cleared for testing extragenital sites, but recommendations for screening extragenital sites prompted many clinical laboratories to validate throat and rectal swabs for use with NAATs, which are more sensitive than culture at these sites.^25,28 The recent FDA approval of extragenital specimen types for 2 commercially available assays may increase the availability of testing for these sites.

Data on the utility of NAATs for detecting chlamydia and gonorrhea in children are limited, and many clinical laboratories have not validated molecular methods for testing in children. Current guidelines specific to this population should be followed regarding test methods and preferred specimen types.^12,29,30

Although gonococcal infection is usually diagnosed with culture-independent molecular methods, antimicrobial resistance is emerging. Thus, failure of the combination of ceftriaxone and azithromycin should prompt culture-based follow-up testing to determine antimicrobial susceptibility.

Strategies for treatment and control

Historically, people treated for gonorrhea have been treated for chlamydia at the same time, as these diseases tend to go together. This can be with a single intramuscular dose of ceftriaxone for the gonorrhea plus a single oral dose of azithromycin for the chlamydia.¹² For patients who have only gonorrhea, this double regimen may help prevent the development of resistant gonorrhea strains.

All the patient’s sexual partners in the previous 60 days should be tested and treated, and expedited partner therapy should be offered if possible. Patients should be advised to have no sexual contact until they complete the treatment, or 7 days after single-dose treatment. Testing should be repeated 3 months after treatment.

A member of the Mycoplasmataceae family, M genitalium was originally identified as a pathogen in the early 1980s but has only recently emerged as an important cause of STI. Studies indicate that it is responsible for 10% to 20% of cases of nongonococcal urethritis and 10% to 30% of cases of cervicitis.^31–33 Additionally, 2% to 22% of cases of pelvic inflammatory disease have evidence of M genitalium.^34,35

However, data on M genitalium prevalence are suspect because the organism is hard to identify—lacking a cell wall, it is undetectable by Gram stain.³⁶ Although it has been isolated in respiratory and synovial fluids, it has so far been recognized to be clinically important only in the urogenital tract. It can persist for years in infected patients by exploiting specialized cell-surface structures to invade cells.³⁶ Once inside a cell, it triggers secretion of mycoplasmal toxins and destructive metabolites such as hydrogen peroxide, evading the host immune system as it does so.³⁷

Testing guidelines for M genitalium

Current guidelines do not recommend routine screening for M genitalium, and no commercial test was available until recently.¹² Although evidence suggests that M genitalium is independently associated with preterm birth and miscarriages,³⁸ routine screening of pregnant women is not recommended.¹²

Testing for M genitalium should be considered in cases of persistent or recurrent nongonococcal urethritis in patients who test negative for gonorrhea and chlamydia or for whom treatment has failed.¹² Many isolates exhibit genotypic resistance to macrolide antibiotics, which are often the first-line therapy for nongonococcal urethritis.³⁹

Further study is needed to evaluate the potential impact of routine screening for M genitalium on the reproductive and sexual health of at-risk populations.

Diagnostic tests for M genitalium

Awareness of M genitalium as a cause of nongonococcal urethritis has been hampered by a dearth of diagnostic tests.⁴⁰ The organism’s fastidious requirements and extremely slow growth preclude culture as a practical method of diagnosis.⁴¹ Serologic assays are dogged by cross-reactivity and poor sensitivity.^42,43 Thus, molecular assays for detecting M genitalium and associated resistance markers are preferred for diagnosis.¹²

Several molecular tests are approved, available, and in use in Europe for diagnosing M genitalium infection,⁴⁰ and in January 2019 the FDA approved a molecular test that can detect M genitalium in urine specimens and vaginal, endocervical, urethral, and penile meatal swabs. Although vaginal swabs are preferred for this assay because they have higher sensitivity (92% for provider-collected and 99% for patient-collected swabs), urine specimens are acceptable, with a sensitivity of 78%.⁴⁴

At least 1 company is seeking FDA clearance for another molecular diagnostic assay for detecting M genitalium and markers of macrolide resistance in urine and genital swab specimens. Such assays may facilitate appropriate treatment.

Clinicians should stay abreast of diagnostic testing options, which are likely to become more readily available soon.

A high rate of macrolide resistance

Because M genitalium lacks a cell wall, antibiotics such as beta-lactams that target cell wall synthesis are ineffective.

Regimens for treating M genitalium are outlined in Table 2.¹² Azithromycin is more effective than doxycycline. However, as many as 50% of strains were macrolide-resistant in a cohort of US female patients.⁴⁵ Given the high incidence of treatment failure with azithromycin 1 g, it is thought that this regimen might select for resistance. For cases in which symptoms persist, a 1- to 2-week course of moxifloxacin is recommended.¹² However, this has not been validated by clinical trials, and failures of the 7-day regimen have been reported.⁴⁶

Partners of patients who test positive for M genitalium should also be tested and undergo clinically applicable screening for nongonococcal urethritis, cervicitis, and pelvic inflammatory disease.¹²

TRICHOMONIASIS

Trichomoniasis, caused by the parasite Trichomonas vaginalis, is the most prevalent nonviral STI in the United States. It disproportionately affects black women, in whom the prevalence is 13%, compared with 1% in non-Hispanic white women.⁴⁷ It is also present in 26% of women with symptoms who are seen in STI clinics and is highly prevalent in incarcerated populations. It is uncommon in men who have sex with men.⁴⁸

In men, trichomoniasis manifests as urethritis, epididymitis, or prostatitis. While most infected women have no symptoms, they may experience vaginitis with discharge that is diffuse, frothy, pruritic, malodorous, or yellow-green. Vaginal and cervical erythema (“strawberry cervix”) can also occur.

Screening for trichomoniasis

Current guidelines of the US Centers for Disease Control and Prevention (CDC) recommend testing for T vaginalis in women who have symptoms and routinely screening in women who are HIV-positive, regardless of symptoms. There is no evidence to support routine screening of pregnant women without symptoms, and pregnant women who do have symptoms should be evaluated according to the same guidelines as for nonpregnant women.¹² Testing can be considered in patients who have no symptoms but who engage in high-risk behaviors and in areas of high prevalence.

A lack of studies using sensitive methods for T vaginalis detection has hampered a true estimation of disease burden and at-risk populations. Screening recommendations may evolve in upcoming clinical guidelines as the field advances.

As infection can recur, women should be retested 3 months after initial diagnosis.¹²

NAAT is the preferred test for trichomoniasis

Commercially available diagnostic tests for trichomoniasis include culture, antigen testing, and NAAT.⁴⁹ While many clinicians do their own wet-mount microscopy for a rapid result, this method has low sensitivity.⁵⁰ Similarly, antigen testing and culture perform poorly compared with NAATs, which are the gold standard for detection.^51,52 A major advantage of NAATs for T vaginalis detection is that they combine high sensitivity and fast results, facilitating diagnosis and appropriate treatment of patients and their partners.

In spite of these benefits, adoption of molecular diagnostic testing for T vaginalis has lagged behind that for chlamydia and gonorrhea.⁵³ FDA-cleared NAATs are available for testing vaginal, cervical, or urine specimens from women, but until recently, there were no approved assays for testing in men. The Cepheid Xpert TV assay, which is valid for male urine specimens to diagnose other sexually transmitted diseases, has demonstrated excellent diagnostic sensitivity for T vaginalis in men and women.⁵⁴ Interestingly, a large proportion of male patients in this study had no symptoms, suggesting that screening of men in high-risk groups may be warranted.

7-day metronidazole treatment beats single-dose treatment

The first-line treatment for trichomoniasis has been a single dose of metronidazole 2 g by mouth, but in a recent randomized controlled trial,⁵⁵ a course of 500 mg by mouth twice a day for 7 days was 45% more effective at 4 weeks than a single dose, and it should now be the preferred regimen.

In clinical trials,⁵⁶ a single dose of tinidazole 2 g orally was equivalent or superior to metronidazole 2 g and had fewer gastrointestinal side effects, but it is more expensive.

Sexually transmitted infections (STIs) such as gonorrhea, chlamydia, and syphilis are still increasing in incidence and probably will continue to do so in the near future. Moreover, drug-resistant strains of Neisseria gonorrhoeae are emerging, as are less-known organisms such as Mycoplasma genitalium.

Now the good news: new tests for STIs are available or are coming! Based on nucleic acid amplification, these tests can be performed at the point of care, so that patients can leave the clinic with an accurate diagnosis and proper treatment for themselves and their sexual partners. Also, the tests can be run on samples collected by the patients themselves, either swabs or urine collections, eliminating the need for invasive sampling and making doctor-shy patients more likely to come in to be treated.¹ We hope that by using these sensitive and accurate tests we can begin to bend the upward curve of STIs and be better antimicrobial stewards.²

This article reviews current issues surrounding STI control, and provides detailed guidance on recognizing, testing for, and treating gonorrhea, chlamydia, trichomoniasis, and M genitalium infection.

STI RATES ARE HIGH AND RISING

STIs are among the most common acute infectious diseases worldwide, with an estimated 1 million new curable cases every day.³ Further, STIs have major impacts on sexual, reproductive, and psychological health.

In the United States, rates of reportable STIs (chlamydia, gonorrhea, and syphilis) are rising.⁴ In addition, more-sensitive tests for trichomoniasis, which is not a reportable infection in any state, have revealed it to be more prevalent than previously thought.⁵

BARRIERS AND CHALLENGES TO DIAGNOSIS

The medical system does not fully meet the needs of some populations, including young people and men who have sex with men, regarding their sexual and reproductive health. 

Ongoing barriers among young people include reluctance to use available health services, limited access to STI testing, worries about confidentiality, and the shame and stigma associated with STIs.⁶

Men who have sex with men have a higher incidence of STIs than other groups. Since STIs are associated with a higher risk of human immunodeficiency virus (HIV) infection, it is important to detect, diagnose, and manage STIs in this group—and in all high-risk groups. Rectal STIs are an independent risk factor for incident HIV infection.⁷ In addition, many men who have sex with men face challenges navigating the emotional, physical, and cognitive aspects of adolescence, a voyage further complicated by mental health issues, unprotected sexual encounters, and substance abuse in many, especially among minority youth.⁸ These same factors also impair their ability to access resources for preventing and treating HIV and other STIs.

STI diagnosis is often missed

Most people who have STIs feel no symptoms, which increases the importance of risk-based screening to detect these infections.^9,10 In many other cases, STIs manifest with nonspecific genitourinary symptoms that are mistaken for urinary tract infection. Tomas et al¹¹ found that of 264 women who presented to an emergency department with genitourinary symptoms or were being treated for urinary tract infection, 175 were given a diagnosis of a urinary tract infection. Of these, 100 (57%) were treated without performing a urine culture; 60 (23%) of the 264 women had 1 or more positive STI tests, 22 (37%) of whom did not receive treatment for an STI.

Poor follow-up of patients and partners

Patients with STIs need to be retested 3 months after treatment to make sure the treatment was effective. Another reason for follow-up is that these patients are at higher risk of another infection within a year.¹²

Although treating patients’ partners has been shown to reduce reinfection rates, fewer than one-third of STIs (including HIV infections) were recognized through partner notification between 2010 and 2012 in a Dutch study, in men who have sex with men and in women.¹³ Challenges included partners who could not be identified among men who have sex with men, failure of heterosexual men to notify their partners, and lower rates of partner notification for HIV.  

In the United States, “expedited partner therapy” allows healthcare providers to provide a prescription or medications to partners of patients diagnosed with chlamydia or gonorrhea without examining the partner.¹⁴ While this approach is legal in most states, implementation can be challenging.¹⁵

STI EVALUATION

History and physical examination

A complete sexual history helps in estimating the patient’s risk of an STI and applying appropriate risk-based screening. Factors such as sexual practices, use of barrier protection, and history of STIs should be discussed.

Physical examination is also important. Although some patients may experience discomfort during a genital or pelvic examination, omitting this step may lead to missed diagnoses in women with STIs.¹⁶

Laboratory testing

Laboratory testing for STIs helps ensure accurate diagnosis and treatment. Empiric treatment without testing could give a patient a false sense of health by missing an infection that is not currently causing symptoms but that could later worsen or have lasting complications. Failure to test patients also misses the opportunity for partner notification, linkage to services, and follow-up testing.

Many of the most common STIs, including gonorrhea, chlamydia, and trichomoniasis, can be detected using vaginal, cervical, or urethral swabs or first-catch urine (from the initial urine stream). In studies that compared various sampling methods,¹⁷ self-collected urine samples for gonorrhea in men were nearly as good as clinician-collected swabs of the urethra. In women, self-collected vaginal swabs for gonorrhea and chlamydia were nearly as good as clinician-collected vaginal swabs. While urine specimens are acceptable for chlamydia testing in women, their sensitivity may be slightly lower than with vaginal and endocervical swab specimens.^18,19

A major advantage of urine specimens for STI testing is that collection is noninvasive and is therefore more likely to be acceptable to patients. Urine testing can also be conducted in a variety of nonclinical settings such as health fairs, pharmacy-based screening programs, and express STI testing sites, thus increasing availability.

To prevent further transmission and morbidity and to aid in public health efforts, it is critical to recognize the cause of infectious cervicitis and urethritis and to screen for STIs according to guidelines.¹² Table 1 summarizes current screening and laboratory testing recommendations.

Gonorrhea and chlamydia are the 2 most frequently reported STIs in the United States, with more than 550,000 cases of gonorrhea and 1.7 million cases of chlamydia reported in 2017.⁴

Both infections present similarly: cervicitis or urethritis characterized by discharge (mucopurulent discharge with gonorrhea) and dysuria. Untreated, they can lead to pelvic inflammatory disease, inflammation, and infertility.

Extragenital infections can be asymptomatic or cause exudative pharyngitis or proctitis. Most people in whom chlamydia is detected from pharyngeal specimens are asymptomatic. When pharyngeal symptoms exist secondary to gonorrheal infection, they typically include sore throat and pharyngeal exudates. However, Komaroff et al,²⁰ in a study of 192 men and women who presented with sore throat, found that only 2 (1%) tested positive for N gonorrhoeae.

Screening for gonorrhea and chlamydia

Best practices include screening for gonorrhea and chlamydia as follows^21–23:

• Every year in sexually active women through age 25 (including during pregnancy) and in older women who have risk factors for infection¹²
• At least every year in men who have sex with men, at all sites of sexual contact (urethra, pharynx, rectum), along with testing for HIV and syphilis
• Every 3 to 6 months in men who have sex with men who have multiple or anonymous partners, who are sexually active and use illicit drugs, or who have partners who use illicit drugs
• Possibly every year in young men who live in high-prevalence areas or who are seen in certain clinical settings, such as STI and adolescent clinics.

Specimens. A vaginal swab is preferred for screening in women. Several studies have shown that self-collected swabs have clinical sensitivity and specificity comparable to that of provider-collected samples.^17,24 First-catch urine or endocervical swabs have similar performance characteristics and are also acceptable. In men, urethral swabs or first-catch urine samples are appropriate for screening for urogenital infections.

Testing methods. Testing for both pathogens should be done simultaneously with a nucleic acid amplification test (NAAT). Commercially available NAATs are more sensitive than culture and antigen testing for detecting gonorrhea and chlamydia.^25–27

Most assays are approved by the US Food and Drug Administration (FDA) for testing vaginal, urethral, cervical, and urine specimens. Until recently, no commercial assay was cleared for testing extragenital sites, but recommendations for screening extragenital sites prompted many clinical laboratories to validate throat and rectal swabs for use with NAATs, which are more sensitive than culture at these sites.^25,28 The recent FDA approval of extragenital specimen types for 2 commercially available assays may increase the availability of testing for these sites.

Data on the utility of NAATs for detecting chlamydia and gonorrhea in children are limited, and many clinical laboratories have not validated molecular methods for testing in children. Current guidelines specific to this population should be followed regarding test methods and preferred specimen types.^12,29,30

Although gonococcal infection is usually diagnosed with culture-independent molecular methods, antimicrobial resistance is emerging. Thus, failure of the combination of ceftriaxone and azithromycin should prompt culture-based follow-up testing to determine antimicrobial susceptibility.

Strategies for treatment and control

Historically, people treated for gonorrhea have been treated for chlamydia at the same time, as these diseases tend to go together. This can be with a single intramuscular dose of ceftriaxone for the gonorrhea plus a single oral dose of azithromycin for the chlamydia.¹² For patients who have only gonorrhea, this double regimen may help prevent the development of resistant gonorrhea strains.

All the patient’s sexual partners in the previous 60 days should be tested and treated, and expedited partner therapy should be offered if possible. Patients should be advised to have no sexual contact until they complete the treatment, or 7 days after single-dose treatment. Testing should be repeated 3 months after treatment.

A member of the Mycoplasmataceae family, M genitalium was originally identified as a pathogen in the early 1980s but has only recently emerged as an important cause of STI. Studies indicate that it is responsible for 10% to 20% of cases of nongonococcal urethritis and 10% to 30% of cases of cervicitis.^31–33 Additionally, 2% to 22% of cases of pelvic inflammatory disease have evidence of M genitalium.^34,35

However, data on M genitalium prevalence are suspect because the organism is hard to identify—lacking a cell wall, it is undetectable by Gram stain.³⁶ Although it has been isolated in respiratory and synovial fluids, it has so far been recognized to be clinically important only in the urogenital tract. It can persist for years in infected patients by exploiting specialized cell-surface structures to invade cells.³⁶ Once inside a cell, it triggers secretion of mycoplasmal toxins and destructive metabolites such as hydrogen peroxide, evading the host immune system as it does so.³⁷

Testing guidelines for M genitalium

Current guidelines do not recommend routine screening for M genitalium, and no commercial test was available until recently.¹² Although evidence suggests that M genitalium is independently associated with preterm birth and miscarriages,³⁸ routine screening of pregnant women is not recommended.¹²

Testing for M genitalium should be considered in cases of persistent or recurrent nongonococcal urethritis in patients who test negative for gonorrhea and chlamydia or for whom treatment has failed.¹² Many isolates exhibit genotypic resistance to macrolide antibiotics, which are often the first-line therapy for nongonococcal urethritis.³⁹

Further study is needed to evaluate the potential impact of routine screening for M genitalium on the reproductive and sexual health of at-risk populations.

Diagnostic tests for M genitalium

Awareness of M genitalium as a cause of nongonococcal urethritis has been hampered by a dearth of diagnostic tests.⁴⁰ The organism’s fastidious requirements and extremely slow growth preclude culture as a practical method of diagnosis.⁴¹ Serologic assays are dogged by cross-reactivity and poor sensitivity.^42,43 Thus, molecular assays for detecting M genitalium and associated resistance markers are preferred for diagnosis.¹²

Several molecular tests are approved, available, and in use in Europe for diagnosing M genitalium infection,⁴⁰ and in January 2019 the FDA approved a molecular test that can detect M genitalium in urine specimens and vaginal, endocervical, urethral, and penile meatal swabs. Although vaginal swabs are preferred for this assay because they have higher sensitivity (92% for provider-collected and 99% for patient-collected swabs), urine specimens are acceptable, with a sensitivity of 78%.⁴⁴

At least 1 company is seeking FDA clearance for another molecular diagnostic assay for detecting M genitalium and markers of macrolide resistance in urine and genital swab specimens. Such assays may facilitate appropriate treatment.

Clinicians should stay abreast of diagnostic testing options, which are likely to become more readily available soon.

A high rate of macrolide resistance

Because M genitalium lacks a cell wall, antibiotics such as beta-lactams that target cell wall synthesis are ineffective.

Regimens for treating M genitalium are outlined in Table 2.¹² Azithromycin is more effective than doxycycline. However, as many as 50% of strains were macrolide-resistant in a cohort of US female patients.⁴⁵ Given the high incidence of treatment failure with azithromycin 1 g, it is thought that this regimen might select for resistance. For cases in which symptoms persist, a 1- to 2-week course of moxifloxacin is recommended.¹² However, this has not been validated by clinical trials, and failures of the 7-day regimen have been reported.⁴⁶

Partners of patients who test positive for M genitalium should also be tested and undergo clinically applicable screening for nongonococcal urethritis, cervicitis, and pelvic inflammatory disease.¹²

TRICHOMONIASIS

Trichomoniasis, caused by the parasite Trichomonas vaginalis, is the most prevalent nonviral STI in the United States. It disproportionately affects black women, in whom the prevalence is 13%, compared with 1% in non-Hispanic white women.⁴⁷ It is also present in 26% of women with symptoms who are seen in STI clinics and is highly prevalent in incarcerated populations. It is uncommon in men who have sex with men.⁴⁸

In men, trichomoniasis manifests as urethritis, epididymitis, or prostatitis. While most infected women have no symptoms, they may experience vaginitis with discharge that is diffuse, frothy, pruritic, malodorous, or yellow-green. Vaginal and cervical erythema (“strawberry cervix”) can also occur.

Screening for trichomoniasis

Current guidelines of the US Centers for Disease Control and Prevention (CDC) recommend testing for T vaginalis in women who have symptoms and routinely screening in women who are HIV-positive, regardless of symptoms. There is no evidence to support routine screening of pregnant women without symptoms, and pregnant women who do have symptoms should be evaluated according to the same guidelines as for nonpregnant women.¹² Testing can be considered in patients who have no symptoms but who engage in high-risk behaviors and in areas of high prevalence.

A lack of studies using sensitive methods for T vaginalis detection has hampered a true estimation of disease burden and at-risk populations. Screening recommendations may evolve in upcoming clinical guidelines as the field advances.

As infection can recur, women should be retested 3 months after initial diagnosis.¹²

NAAT is the preferred test for trichomoniasis

Commercially available diagnostic tests for trichomoniasis include culture, antigen testing, and NAAT.⁴⁹ While many clinicians do their own wet-mount microscopy for a rapid result, this method has low sensitivity.⁵⁰ Similarly, antigen testing and culture perform poorly compared with NAATs, which are the gold standard for detection.^51,52 A major advantage of NAATs for T vaginalis detection is that they combine high sensitivity and fast results, facilitating diagnosis and appropriate treatment of patients and their partners.

In spite of these benefits, adoption of molecular diagnostic testing for T vaginalis has lagged behind that for chlamydia and gonorrhea.⁵³ FDA-cleared NAATs are available for testing vaginal, cervical, or urine specimens from women, but until recently, there were no approved assays for testing in men. The Cepheid Xpert TV assay, which is valid for male urine specimens to diagnose other sexually transmitted diseases, has demonstrated excellent diagnostic sensitivity for T vaginalis in men and women.⁵⁴ Interestingly, a large proportion of male patients in this study had no symptoms, suggesting that screening of men in high-risk groups may be warranted.

7-day metronidazole treatment beats single-dose treatment

The first-line treatment for trichomoniasis has been a single dose of metronidazole 2 g by mouth, but in a recent randomized controlled trial,⁵⁵ a course of 500 mg by mouth twice a day for 7 days was 45% more effective at 4 weeks than a single dose, and it should now be the preferred regimen.

In clinical trials,⁵⁶ a single dose of tinidazole 2 g orally was equivalent or superior to metronidazole 2 g and had fewer gastrointestinal side effects, but it is more expensive.

All cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus.¹ But most cases of Barrett esophagus go undiagnosed. And Barrett esophagus is seen in 5% to 15% of patients with gastroesophageal reflux disease.² These facts clearly emphasize the need for screening. Here, we review the rationale and recommendations for screening and surveillance, as well as the range of treatment options.

SCOPE OF THE PROBLEM

The American Cancer Society estimated there were 17,290 new cases of esophageal cancer and 15,850 deaths from it in the United States in 2018.³ Of the 2 main histologic types of esophageal cancer, adenocarcinoma and squamous cell cancer, adenocarcinoma is more common in the United States.

The precursor lesion is Barrett esophagus, defined as an extension of salmon-colored mucosa at least 1 cm into the tubular esophagus proximal to the gastroesophageal junction, with biopsy confirmation of intestinal metaplasia.⁴

The natural course of progression to dysplasia and cancer in Barrett esophagus is unknown but is thought to be stepwise, from no dysplasia to low-grade dysplasia to high-grade dysplasia and cancer, and the cancer risk depends on the degree of dysplasia: the annual risk is 0.33% if there is no dysplasia, 0.54% with low-grade dysplasia, and 7% with high-grade dysplasia.⁴

Although all cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus,¹ more than 90% of patients with newly diagnosed esophageal adenocarcinoma do not have a prior diagnosis of Barrett esophagus.⁵ Therefore, there is a substantial unmet need to expand screening for Barrett esophagus in people at risk.

GASTROESOPHAGEAL REFLUX DISEASE IS A RISK FACTOR FOR CANCER

The rationale behind screening is that detecting Barrett esophagus early and intervening in a timely manner in patients at higher risk of developing adenocarcinoma will decrease mortality.

Chronic gastroesophageal reflux disease is a strong risk factor for esophageal adenocarcinoma (odds ratio [OR] 7.7, 95% confidence interval [CI] 5.3–11.4), and the risk increases when symptoms are long-standing (> 20 years) or severe (OR 43.5, 95% CI 18.3–103.5) or occur daily (OR 5.5, 95% CI 3.2–9.3).⁶

Reflux symptoms are scored as follows:

• Heartburn only, 1 point
• Regurgitation only, 1 point
• Heartburn with regurgitation, 1.5 points
• Nightly symptoms (2 points if yes, 0 if no)
• Symptoms once a week, 0 points; 2 to 6 times a week, 1 point; 7 to 15 times a week, 2 points; more than 15 times a week, 3 points.⁶

A score of 4.5 or higher indicates severe reflux disease. However, it is worth noting that the annual incidence of esophageal adenocarcinoma in patients with long-term gastroesophageal reflux disease is less than 0.001%.⁷

RISK FACTORS FOR BARRETT ESOPHAGUS

Risk factors for Barrett esophagus include:

Male sex. Barrett esophagus is more prevalent in men than in women, at a ratio of 2 to 1; but in individuals under age 50, the ratio is 4 to 1.⁸

Age 50 or older. Barrett esophagus is usually diagnosed in the sixth to seventh decade of life, and the prevalence increases from 2.1% in the third decade to 9.3% in the sixth decade.⁹

White race. It is more prevalent in whites than in blacks (5.0% vs 1.5%, P < .0001).¹⁰

Central obesity. Waist circumference is an independent risk factor: every 5-cm increase carries an OR of 1.14 (95% CI 1.03–1.27, P = .02).¹¹

Cigarette smoking increases the risk of Barrett esophagus (OR 1.42; 95% CI 1.15–1.76).¹²

A family history of Barrett esophagus or esophageal adenocarcinoma is a strong risk factor (OR 12, 95% CI 3.3–44.8). In 1 study, the risk in first- and second-degree relatives of patients with Barrett esophagus was 24%, compared with 5% in a control population (P < .005).¹³

SCREENING GUIDELINES AND DRAWBACKS

The standard screening method is esophagogastroduodenoscopy with sedation, with careful visual inspection and 4-quadrant biopsies every 2 cm using the Seattle protocol, ie, including biopsy of any mucosal irregularities in salmon-colored mucosa above the gastroesophageal junction (Figure 1).⁴

Endoscopic screening is cost-effective, costing $10,440 per quality-adjusted life-year saved, which is well below the accepted threshold of less than $100,000.¹⁴ However, it is still expensive, invasive, and not ideal for screening large populations.

Less-invasive methods under study

Less-invasive, less-expensive methods being tested for mass screening include:

Unsedated transnasal endoscopy. Done with only topical anesthesia, it has high diagnostic accuracy and is quicker and more cost-effective than standard esophagogastroduodenoscopy, with fewer adverse effects. However, the procedure has not yet gained widespread acceptance for regular use by gastroenterologists.¹⁵

A swallowable sponge. Another promising test is cell collection using the Cytosponge Cell Collection Device (Medtronic, Minneapolis, MN). An encapsulated compressed sponge with a string attached is swallowed; in the stomach, the capsule dissolves, and the sponge expands and is then withdrawn using the attached string. The obtained cytology sample from the lower esophagus is then tested for trefoil factor 3, a protein biomarker for Barrett esophagus.¹⁶

A retractable balloon. The EsoCheck Cell Collection Device is a retractable balloon attached to a string. When swallowed, it gathers distal esophageal cells for detecting methylated DNA markers for Barrett esophagus.¹⁷

Esophageal capsule endoscopy uses a camera to visualize the esophagus, but lacks the ability to obtain biopsy samples.

Other screening methods are being tested, although data are limited. Liquid biopsy uses a blood sample to detect microRNAs that are dysregulated in cancer. The “electronic nose” is a device that detects exhaled volatile organic compounds altered in Barrett esophagus. Another test involves taking an oral wash sample to study the oral microbiome for a pattern associated with adenocarcinoma.^18–21

Surveillance in Barrett esophagus aims to detect premalignant changes or early-stage adenocarcinoma to provide longer survival and lower cancer-related mortality. Recent evidence suggests that patients with esophageal adenocarcinoma that is diagnosed in a Barrett esophagus surveillance program have an earlier stage of disease and therefore a survival benefit.²²

Patient education is essential

Before enrolling a patient in a surveillance program, the clinician should explain the risks, benefits, and limitations, the importance of periodic endoscopy, and the possible eventual need for endoscopic therapy or surgery.

The endoscopic procedure

Surveillance involves examination by high-definition white-light endoscopy, with random 4-quadrant biopsies every 2 cm (or every 1 cm in patients with a history of dysplasia) and biopsy of any mucosal irregularity (nodule, ulcer, or other visible lesion). The degree of dysplasia determines the frequency of follow-up surveillance intervals and the need for endoscopic eradication therapy, as presented in professional society guidelines (Table 1).^4,23,24

Advanced methods for detecting dysplasia

Newer endoscopic surveillance techniques include dye-based chromoendoscopy, narrow-band imaging, confocal laser endomicroscopy, volumetric laser endomicroscopy, and wide-area transepithelial sampling with computer-assisted 3-dimensional analysis. All these techniques are used to increase the detection of dysplasia. Of these, dye-based chromoendoscopy, narrow-band imaging, and confocal laser endomicroscopy meet current criteria of the American Society for Gastrointestinal Endoscopy for preservation and incorporation of valuable endoscopic innovations.²³

MANAGEMENT OF NONDYSPLASTIC BARRETT ESOPHAGUS

A proton pump inhibitor (PPI) is recommended to control reflux symptoms in patients with nondysplastic Barrett esophagus. But it is important to counsel patients on additional ways to protect against esophageal adenocarcinoma, such as:

• Low to moderate alcohol consumption
• Regular physical activity
• Increased dietary intake of fruits, vegetables, folate, fiber, beta-carotene, and vitamin C
• Weight control
• Smoking cessation.²⁵

Surveillance endoscopy with 4-quadrant biopsies at 2-cm intervals is recommended every 3 to 5 years (Table 1).

DOES CHEMOPREVENTION HAVE A ROLE?

Chemoprevention is an exciting area of research in preventing progression to adenocarcinoma in patients with Barrett esophagus. Various drugs such as aspirin, other nonsteroidal anti-inflammatory drugs (NSAIDs), PPIs, metformin, and statins have been studied.

Aspirin

Aspirin has been shown to prevent development of Barrett esophagus in patients with reflux disease,²⁶ but more studies are needed to validate those findings.

PPIs

Gastroesophageal reflux disease is a primary risk factor for esophageal adenocarcinoma, and gastric acid suppression with PPIs reduces cancer risk. PPI therapy is associated with a 71% decrease in the risk of high-grade dysplasia and adenocarcinoma in patients with Barrett esophagus (OR 0.29, 95% CI 0.12–0.79).²⁷ Long-term therapy (> 2 to 3 years) has a higher protective effect (adjusted OR 0.45, 95% CI 0.19–1.06) than short-term therapy (< 2 to 3 years) (adjusted OR 1.09, 95% CI 0.47–2.56).²⁷

NSAIDs

NSAIDs, including aspirin, have been associated with decreased risk of colon, stomach, lung, breast, and esophageal cancer due to their potential to inhibit cyclooxygenase 2 (COX-2) enzymes.

A meta-analysis demonstrated that aspirin and NSAIDs led to a 32% reduction in the risk of adenocarcinoma (OR 0.68, 95% CI 0.56–0.83). The benefit was even greater if the drug was taken daily or more frequently (OR 0.56, 95% CI 0.43–0.73, P < .001) or was taken for 10 or more years (OR 0.63, 95% CI 0.45–0.90, P = .04).²⁸

PPI plus aspirin

The best evidence for the role of PPIs and aspirin in reducing the risk of dysplasia comes from the Aspirin and Esomeprazole Chemoprevention in Barrett’s Metaplasia Trial.²⁹ This randomized, controlled trial compared 4 regimens consisting of esomeprazole (a PPI) in either a high dose (40 mg twice daily) or a low dose (20 mg once daily) plus either aspirin (300 or 320 mg per day) or no aspirin in 2,557 patients with Barrett esophagus. The composite end point was the time to all-cause mortality, adenocarcinoma, or high-grade dysplasia.

At a median follow-up of 8.9 years, the combination of high-dose esomeprazole plus aspirin had the strongest effect compared with low-dose esomeprazole without aspirin (time ratio 1.59, 95% CI 1.14–2.23, P = .0068). The number needed to treat was 34 for esomeprazole and 43 for aspirin.²⁹

Based on these data, we can conclude that aspirin and PPIs can prevent dysplasia and all-cause mortality in Barrett esophagus.

Metformin: No evidence of benefit

Metformin was studied as a protective agent against obesity-associated cancers including esophageal adenocarcinoma, as it reduces insulin levels.

In a randomized controlled trial³⁰ in 74 patients with Barrett esophagus, metformin (starting at 500 mg daily, increasing to 2,000 mg/day by week 4) was compared with placebo. At 12 weeks, the percent change in esophageal levels of the biomarker pS6K1—an intracellular mediator of insulin and insulin-like growth factor activation in Barrett epithelium—did not differ significantly between the 2 groups (1.4% with metformin vs −14.7% with placebo; 1-sided P = .80). This suggested that metformin did not significantly alter proliferation or apoptosis in Barrett epithelium, despite reducing serum insulin levels and insulin resistance. Thus, metformin did not demonstrate a chemoprotective effect in preventing the progression of Barrett esophagus to adenocarcinoma.

Vitamin D affects genes regulating proliferation, apoptosis, and differentiation, and has therefore been studied as a potential antineoplastic agent. Its deficiency has also been associated with increased risk of esophageal adenocarcinoma. However, its efficacy in chemoprevention is unclear.³¹

One study found no association between serum 25-hydroxyvitamin D levels and prevalence of dysplasia in Barrett esophagus (P = .90). An increase in vitamin D levels had no effect on progression to dysplasia or cancer (for every 5-nmol/L increase from baseline, hazard ratio 0.98, P = .62).³²

In another study, supplementation with vitamin D₃ (cholecalciferol 50,000 IU weekly) plus a PPI for 12 weeks significantly improved the serum 25-hydroxyvitamin D levels without significant changes in gene expression from Barrett epithelium.³³ These findings were confirmed in a meta-analysis that showed no consistent association between vitamin D exposure and risk of esophageal neoplasm.³⁴

Thus, there is currently no evidence to support vitamin D for chemoprevention in Barrett esophagus or esophageal adenocarcinoma.

Statins

In addition to lowering cholesterol, statins have antiproliferative, pro-apoptotic, anti-angiogenic, and immunomodulatory effects that prevent cancer, leading to a 41% reduction in the risk of adenocarcinoma in patients with Barrett esophagus in one study (adjusted OR 0.59, 95% CI 0.45–0.78); the number needed to treat with statins to prevent 1 case of adenocarcinoma was 389.³⁵

A meta-analysis also showed that statin use was associated with a lower risk of progression of Barrett esophagus (OR 0.48, 95% CI 0.31–0.73).³⁶

In general, statins appear promising for chemoprevention, but more study is needed.

When is chemoprevention appropriate?

Chemoprevention is not recommended for all patients with Barrett esophagus, given that the condition affects 1% to 2% of the US adult population, and very few patients have progression to esophageal adenocarcinoma. Rather, chemoprevention may be considered in patients with Barrett esophagus and multiple risk factors for adenocarcinoma.

INDEFINITE DYSPLASIA

In Barrett esophagus with indefinite dysplasia, either the epithelial abnormalities are insufficient for a diagnosis of dysplasia, or the nature of the epithelial abnormalities is uncertain due to inflammation or technical difficulties with specimen processing. The risk of high-grade dysplasia or cancer within 1 year of the diagnosis of indefinite dysplasia varies between 1.9% and 15%.³⁷ The recommendation for management is to optimize acid-suppressive therapy for 3 to 6 months and then to repeat esophagogastroduodenoscopy. If indefinite dysplasia is noted again, repeat endoscopy in 12 months is recommended.²

ENDOSCOPIC ERADICATION: AN OVERVIEW

Because dysplasia in Barrett esophagus carries a high risk of progression to cancer, the standard of care is endoscopic mucosal resection of visible lesions, followed by ablation of the flat mucosa, with the aim of achieving complete eradication of intestinal metaplasia.^4,38 The initial endoscopic treatment is followed by outpatient sessions every 8 to 10 weeks until the dysplasia is eradicated. A key part of treatment during this time is maximal acid suppression with a PPI twice daily and a histamine-2 blocker at night. In rare cases, fundoplication is required to control reflux refractory to medical therapy.

After eradication is confirmed, continued surveillance is necessary, as recurrences have been reported at a rate of 4.8% per year for intestinal metaplasia, and 2% per year for dysplasia.³⁹

Current endoscopic resection techniques

Endoscopic resection techniques include mucosal resection, submucosal dissection, radio­frequency ablation, cryotherapy, argon plasma coagulation, and photodynamic therapy (Figure 2).

In mucosal resection, the lesion is either suctioned into a band ligator, after which a band is placed around the lesion, or suctioned into a cap fitted at the end of the endoscope, after which the lesion is removed using a snare.

In submucosal dissection, a liquid is injected into the submucosa to lift the lesion, making it easier to remove. The procedure is technically complex and requires additional training.

In radiofrequency ablation, a special catheter is passed through the endoscope to ablate the affected epithelium by thermal injury. Argon plasma coagulation works in a similar way, but uses ionized argon gas to induce thermal coagulation of metaplastic epithelium.

Cryotherapy produces cellular injury by rapid freezing and thawing of tissue using a cryogen such as liquid nitrogen or nitrous oxide.

In photodynamic therapy, a photosensitizer (porfimer sodium) is administered and taken up preferentially by metaplastic epithelium; it is then activated by transmission of red light using the endoscope, leading to destruction of the metaplastic epithelium.

Of the different techniques, radiofrequency ablation has the most evidence for efficacy and hence is the most commonly used.

All of these procedures are generally well tolerated and have favorable side-effect profiles. After radiofrequency ablation with or without mucosal resection, esophageal strictures are noted in 5.6% of patients, and bleeding and perforation occur rarely (1% and 0.6% of patients, respectively).⁴⁰ Submucosal dissection is associated with a higher rate of complications such as stricture formation (11% of patients) and bleeding or perforation (1.5% of patients).⁴¹

Most patients with low-grade dysplasia (73%) are down-staged to nondysplastic Barrett esophagus or to indefinite for dysplasia after review by expert pathologists.⁴² Patients with confirmed and persistent low-grade dysplasia are at higher risk of progression.⁴³

Once low-grade dysplasia is confirmed by a second gastrointestinal pathologist, the patient should undergo endoscopic ablation. A landmark study by Shaheen et al⁴⁴ demonstrated the benefit of radiofrequency ablation in achieving complete eradication of dysplasia (90.5% vs 22.7% for a sham procedure) and complete eradication of intestinal metaplasia (77.4% vs 2.3% for a sham procedure). In another trial of 136 patients with low-grade dysplasia followed for 3 years, Phoa et al⁴⁵ demonstrated that radiofrequency ablation reduced the rate of progression to high-grade dysplasia by 25% and to adenocarcinoma by 7.4% compared with endoscopic surveillance.

Patients with confirmed low-grade dysplasia who do not undergo eradication therapy should have surveillance endoscopy every 6 to 12 months (Table 1).

HIGH-GRADE DYSPLASIA: RECOMMENDED MANAGEMENT

As with low-grade dysplasia, the diagnosis of high-grade dysplasia needs to be confirmed by a second pathologist with gastrointestinal expertise. In the past, the treatment was esophagectomy, but due to lower morbidity and equivalent efficacy of radiofrequency ablation,⁴⁶ the current treatment of choice is endoscopic mucosal resection of raised lesions, followed by radiofrequency ablation of the entire affected segment.

In the study by Shaheen et al,⁴⁴ 42 patients with high-grade dysplasia were randomized to radiofrequency ablation and 21 to a sham procedure, and 81% of ablation patients achieved complete eradication of dysplasia vs 19% with the sham procedure. Eradication of intestinal metaplasia was achieved in 77% of ablation patients vs 2% of patients with the sham therapy. Results of 3-year follow-up from the same cohort showed complete eradication of dysplasia in 98% and of intestinal metaplasia in 91%.⁴⁷

Endoscopic eradication therapy is recommended for all patients with Barrett esophagus and high-grade dysplasia without a life-limiting comorbidity. Alternatively, surveillance every 3 months is an option if the patient does not wish to undergo eradication therapy. Radiofrequency ablation is more cost-effective than esophagectomy or endoscopic surveillance followed by treatment once patients develop adenocarcinoma.^48,49

EARLY ESOPHAGEAL ADENOCARCINOMA: RECOMMENDED MANAGEMENT

Adenocarcinoma limited to the mucosa and without evidence of nodal involvement can be resected endoscopically. In patients with localized cancer, mucosal resection is done not only for therapeutic purposes but also for staging. Ideal management is multidisciplinary, including a gastroenterologist, thoracic surgeon, oncologist, pathologist, and radiation oncologist.

If lesions have features suggesting submucosal invasion or are greater than 1.5 cm in size, or if it is difficult to separate (ie, lift) the mucosa from the submucosal layer with injection of saline, then submucosal dissection is recommended.⁵⁰ Because of the risk of metachronous lesions, ablation of the remaining Barrett esophagus mucosa is recommended after resection of cancer.

Endoscopic eradication is highly effective and durable for the treatment of intramucosal esophageal adenocarcinoma. In a study of 1,000 patients, 963 patients (96.3%) had achieved a complete response; 12 patients (3.7%) underwent surgery after eradication failed during a follow-up of almost 5 years.⁵¹ Metachronous lesions or recurrence of cancer developed during the follow-up period in 140 patients (14.5%) but were successfully treated endoscopically in 115, resulting in a long-term complete remission rate of 93.8%.

POSTABLATION MANAGEMENT

Because of the risk of recurrence of dysplasia after ablation, long-term PPI therapy and surveillance are recommended.

Surveillance endoscopy involves 4-quadrant biopsies taken every 1 cm from the entire length of segment where Barrett esophagus had been seen before ablation.

The timing of surveillance intervals depends on the preablation grade of dysplasia. For low-grade dysplasia, the recommendation is every 6 months for the first year after ablation and, if there is no recurrence of dysplasia, annually after that.² After treatment of high-grade dysplasia or intramucosal adenocarcinoma, the recommendation is every 3 months for the first year, every 6 months in the second year, and then annually.²

All cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus.¹ But most cases of Barrett esophagus go undiagnosed. And Barrett esophagus is seen in 5% to 15% of patients with gastroesophageal reflux disease.² These facts clearly emphasize the need for screening. Here, we review the rationale and recommendations for screening and surveillance, as well as the range of treatment options.

SCOPE OF THE PROBLEM

The American Cancer Society estimated there were 17,290 new cases of esophageal cancer and 15,850 deaths from it in the United States in 2018.³ Of the 2 main histologic types of esophageal cancer, adenocarcinoma and squamous cell cancer, adenocarcinoma is more common in the United States.

The precursor lesion is Barrett esophagus, defined as an extension of salmon-colored mucosa at least 1 cm into the tubular esophagus proximal to the gastroesophageal junction, with biopsy confirmation of intestinal metaplasia.⁴

The natural course of progression to dysplasia and cancer in Barrett esophagus is unknown but is thought to be stepwise, from no dysplasia to low-grade dysplasia to high-grade dysplasia and cancer, and the cancer risk depends on the degree of dysplasia: the annual risk is 0.33% if there is no dysplasia, 0.54% with low-grade dysplasia, and 7% with high-grade dysplasia.⁴

Although all cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus,¹ more than 90% of patients with newly diagnosed esophageal adenocarcinoma do not have a prior diagnosis of Barrett esophagus.⁵ Therefore, there is a substantial unmet need to expand screening for Barrett esophagus in people at risk.

GASTROESOPHAGEAL REFLUX DISEASE IS A RISK FACTOR FOR CANCER

The rationale behind screening is that detecting Barrett esophagus early and intervening in a timely manner in patients at higher risk of developing adenocarcinoma will decrease mortality.

Chronic gastroesophageal reflux disease is a strong risk factor for esophageal adenocarcinoma (odds ratio [OR] 7.7, 95% confidence interval [CI] 5.3–11.4), and the risk increases when symptoms are long-standing (> 20 years) or severe (OR 43.5, 95% CI 18.3–103.5) or occur daily (OR 5.5, 95% CI 3.2–9.3).⁶

Reflux symptoms are scored as follows:

• Heartburn only, 1 point
• Regurgitation only, 1 point
• Heartburn with regurgitation, 1.5 points
• Nightly symptoms (2 points if yes, 0 if no)
• Symptoms once a week, 0 points; 2 to 6 times a week, 1 point; 7 to 15 times a week, 2 points; more than 15 times a week, 3 points.⁶

A score of 4.5 or higher indicates severe reflux disease. However, it is worth noting that the annual incidence of esophageal adenocarcinoma in patients with long-term gastroesophageal reflux disease is less than 0.001%.⁷

RISK FACTORS FOR BARRETT ESOPHAGUS

Risk factors for Barrett esophagus include:

Male sex. Barrett esophagus is more prevalent in men than in women, at a ratio of 2 to 1; but in individuals under age 50, the ratio is 4 to 1.⁸

Age 50 or older. Barrett esophagus is usually diagnosed in the sixth to seventh decade of life, and the prevalence increases from 2.1% in the third decade to 9.3% in the sixth decade.⁹

White race. It is more prevalent in whites than in blacks (5.0% vs 1.5%, P < .0001).¹⁰

Central obesity. Waist circumference is an independent risk factor: every 5-cm increase carries an OR of 1.14 (95% CI 1.03–1.27, P = .02).¹¹

Cigarette smoking increases the risk of Barrett esophagus (OR 1.42; 95% CI 1.15–1.76).¹²

A family history of Barrett esophagus or esophageal adenocarcinoma is a strong risk factor (OR 12, 95% CI 3.3–44.8). In 1 study, the risk in first- and second-degree relatives of patients with Barrett esophagus was 24%, compared with 5% in a control population (P < .005).¹³

SCREENING GUIDELINES AND DRAWBACKS

The standard screening method is esophagogastroduodenoscopy with sedation, with careful visual inspection and 4-quadrant biopsies every 2 cm using the Seattle protocol, ie, including biopsy of any mucosal irregularities in salmon-colored mucosa above the gastroesophageal junction (Figure 1).⁴

Endoscopic screening is cost-effective, costing $10,440 per quality-adjusted life-year saved, which is well below the accepted threshold of less than $100,000.¹⁴ However, it is still expensive, invasive, and not ideal for screening large populations.

Less-invasive methods under study

Less-invasive, less-expensive methods being tested for mass screening include:

Unsedated transnasal endoscopy. Done with only topical anesthesia, it has high diagnostic accuracy and is quicker and more cost-effective than standard esophagogastroduodenoscopy, with fewer adverse effects. However, the procedure has not yet gained widespread acceptance for regular use by gastroenterologists.¹⁵

A swallowable sponge. Another promising test is cell collection using the Cytosponge Cell Collection Device (Medtronic, Minneapolis, MN). An encapsulated compressed sponge with a string attached is swallowed; in the stomach, the capsule dissolves, and the sponge expands and is then withdrawn using the attached string. The obtained cytology sample from the lower esophagus is then tested for trefoil factor 3, a protein biomarker for Barrett esophagus.¹⁶

A retractable balloon. The EsoCheck Cell Collection Device is a retractable balloon attached to a string. When swallowed, it gathers distal esophageal cells for detecting methylated DNA markers for Barrett esophagus.¹⁷

Esophageal capsule endoscopy uses a camera to visualize the esophagus, but lacks the ability to obtain biopsy samples.

Other screening methods are being tested, although data are limited. Liquid biopsy uses a blood sample to detect microRNAs that are dysregulated in cancer. The “electronic nose” is a device that detects exhaled volatile organic compounds altered in Barrett esophagus. Another test involves taking an oral wash sample to study the oral microbiome for a pattern associated with adenocarcinoma.^18–21

Surveillance in Barrett esophagus aims to detect premalignant changes or early-stage adenocarcinoma to provide longer survival and lower cancer-related mortality. Recent evidence suggests that patients with esophageal adenocarcinoma that is diagnosed in a Barrett esophagus surveillance program have an earlier stage of disease and therefore a survival benefit.²²

Patient education is essential

Before enrolling a patient in a surveillance program, the clinician should explain the risks, benefits, and limitations, the importance of periodic endoscopy, and the possible eventual need for endoscopic therapy or surgery.

The endoscopic procedure

Surveillance involves examination by high-definition white-light endoscopy, with random 4-quadrant biopsies every 2 cm (or every 1 cm in patients with a history of dysplasia) and biopsy of any mucosal irregularity (nodule, ulcer, or other visible lesion). The degree of dysplasia determines the frequency of follow-up surveillance intervals and the need for endoscopic eradication therapy, as presented in professional society guidelines (Table 1).^4,23,24

Advanced methods for detecting dysplasia

Newer endoscopic surveillance techniques include dye-based chromoendoscopy, narrow-band imaging, confocal laser endomicroscopy, volumetric laser endomicroscopy, and wide-area transepithelial sampling with computer-assisted 3-dimensional analysis. All these techniques are used to increase the detection of dysplasia. Of these, dye-based chromoendoscopy, narrow-band imaging, and confocal laser endomicroscopy meet current criteria of the American Society for Gastrointestinal Endoscopy for preservation and incorporation of valuable endoscopic innovations.²³

MANAGEMENT OF NONDYSPLASTIC BARRETT ESOPHAGUS

A proton pump inhibitor (PPI) is recommended to control reflux symptoms in patients with nondysplastic Barrett esophagus. But it is important to counsel patients on additional ways to protect against esophageal adenocarcinoma, such as:

• Low to moderate alcohol consumption
• Regular physical activity
• Increased dietary intake of fruits, vegetables, folate, fiber, beta-carotene, and vitamin C
• Weight control
• Smoking cessation.²⁵

Surveillance endoscopy with 4-quadrant biopsies at 2-cm intervals is recommended every 3 to 5 years (Table 1).

DOES CHEMOPREVENTION HAVE A ROLE?

Chemoprevention is an exciting area of research in preventing progression to adenocarcinoma in patients with Barrett esophagus. Various drugs such as aspirin, other nonsteroidal anti-inflammatory drugs (NSAIDs), PPIs, metformin, and statins have been studied.

Aspirin

Aspirin has been shown to prevent development of Barrett esophagus in patients with reflux disease,²⁶ but more studies are needed to validate those findings.

PPIs

Gastroesophageal reflux disease is a primary risk factor for esophageal adenocarcinoma, and gastric acid suppression with PPIs reduces cancer risk. PPI therapy is associated with a 71% decrease in the risk of high-grade dysplasia and adenocarcinoma in patients with Barrett esophagus (OR 0.29, 95% CI 0.12–0.79).²⁷ Long-term therapy (> 2 to 3 years) has a higher protective effect (adjusted OR 0.45, 95% CI 0.19–1.06) than short-term therapy (< 2 to 3 years) (adjusted OR 1.09, 95% CI 0.47–2.56).²⁷

NSAIDs

NSAIDs, including aspirin, have been associated with decreased risk of colon, stomach, lung, breast, and esophageal cancer due to their potential to inhibit cyclooxygenase 2 (COX-2) enzymes.

A meta-analysis demonstrated that aspirin and NSAIDs led to a 32% reduction in the risk of adenocarcinoma (OR 0.68, 95% CI 0.56–0.83). The benefit was even greater if the drug was taken daily or more frequently (OR 0.56, 95% CI 0.43–0.73, P < .001) or was taken for 10 or more years (OR 0.63, 95% CI 0.45–0.90, P = .04).²⁸

PPI plus aspirin

The best evidence for the role of PPIs and aspirin in reducing the risk of dysplasia comes from the Aspirin and Esomeprazole Chemoprevention in Barrett’s Metaplasia Trial.²⁹ This randomized, controlled trial compared 4 regimens consisting of esomeprazole (a PPI) in either a high dose (40 mg twice daily) or a low dose (20 mg once daily) plus either aspirin (300 or 320 mg per day) or no aspirin in 2,557 patients with Barrett esophagus. The composite end point was the time to all-cause mortality, adenocarcinoma, or high-grade dysplasia.

At a median follow-up of 8.9 years, the combination of high-dose esomeprazole plus aspirin had the strongest effect compared with low-dose esomeprazole without aspirin (time ratio 1.59, 95% CI 1.14–2.23, P = .0068). The number needed to treat was 34 for esomeprazole and 43 for aspirin.²⁹

Based on these data, we can conclude that aspirin and PPIs can prevent dysplasia and all-cause mortality in Barrett esophagus.

Metformin: No evidence of benefit

Metformin was studied as a protective agent against obesity-associated cancers including esophageal adenocarcinoma, as it reduces insulin levels.

In a randomized controlled trial³⁰ in 74 patients with Barrett esophagus, metformin (starting at 500 mg daily, increasing to 2,000 mg/day by week 4) was compared with placebo. At 12 weeks, the percent change in esophageal levels of the biomarker pS6K1—an intracellular mediator of insulin and insulin-like growth factor activation in Barrett epithelium—did not differ significantly between the 2 groups (1.4% with metformin vs −14.7% with placebo; 1-sided P = .80). This suggested that metformin did not significantly alter proliferation or apoptosis in Barrett epithelium, despite reducing serum insulin levels and insulin resistance. Thus, metformin did not demonstrate a chemoprotective effect in preventing the progression of Barrett esophagus to adenocarcinoma.

Vitamin D affects genes regulating proliferation, apoptosis, and differentiation, and has therefore been studied as a potential antineoplastic agent. Its deficiency has also been associated with increased risk of esophageal adenocarcinoma. However, its efficacy in chemoprevention is unclear.³¹

One study found no association between serum 25-hydroxyvitamin D levels and prevalence of dysplasia in Barrett esophagus (P = .90). An increase in vitamin D levels had no effect on progression to dysplasia or cancer (for every 5-nmol/L increase from baseline, hazard ratio 0.98, P = .62).³²

In another study, supplementation with vitamin D₃ (cholecalciferol 50,000 IU weekly) plus a PPI for 12 weeks significantly improved the serum 25-hydroxyvitamin D levels without significant changes in gene expression from Barrett epithelium.³³ These findings were confirmed in a meta-analysis that showed no consistent association between vitamin D exposure and risk of esophageal neoplasm.³⁴

Thus, there is currently no evidence to support vitamin D for chemoprevention in Barrett esophagus or esophageal adenocarcinoma.

Statins

In addition to lowering cholesterol, statins have antiproliferative, pro-apoptotic, anti-angiogenic, and immunomodulatory effects that prevent cancer, leading to a 41% reduction in the risk of adenocarcinoma in patients with Barrett esophagus in one study (adjusted OR 0.59, 95% CI 0.45–0.78); the number needed to treat with statins to prevent 1 case of adenocarcinoma was 389.³⁵

A meta-analysis also showed that statin use was associated with a lower risk of progression of Barrett esophagus (OR 0.48, 95% CI 0.31–0.73).³⁶

In general, statins appear promising for chemoprevention, but more study is needed.

When is chemoprevention appropriate?

Chemoprevention is not recommended for all patients with Barrett esophagus, given that the condition affects 1% to 2% of the US adult population, and very few patients have progression to esophageal adenocarcinoma. Rather, chemoprevention may be considered in patients with Barrett esophagus and multiple risk factors for adenocarcinoma.

INDEFINITE DYSPLASIA

In Barrett esophagus with indefinite dysplasia, either the epithelial abnormalities are insufficient for a diagnosis of dysplasia, or the nature of the epithelial abnormalities is uncertain due to inflammation or technical difficulties with specimen processing. The risk of high-grade dysplasia or cancer within 1 year of the diagnosis of indefinite dysplasia varies between 1.9% and 15%.³⁷ The recommendation for management is to optimize acid-suppressive therapy for 3 to 6 months and then to repeat esophagogastroduodenoscopy. If indefinite dysplasia is noted again, repeat endoscopy in 12 months is recommended.²

ENDOSCOPIC ERADICATION: AN OVERVIEW

Because dysplasia in Barrett esophagus carries a high risk of progression to cancer, the standard of care is endoscopic mucosal resection of visible lesions, followed by ablation of the flat mucosa, with the aim of achieving complete eradication of intestinal metaplasia.^4,38 The initial endoscopic treatment is followed by outpatient sessions every 8 to 10 weeks until the dysplasia is eradicated. A key part of treatment during this time is maximal acid suppression with a PPI twice daily and a histamine-2 blocker at night. In rare cases, fundoplication is required to control reflux refractory to medical therapy.

After eradication is confirmed, continued surveillance is necessary, as recurrences have been reported at a rate of 4.8% per year for intestinal metaplasia, and 2% per year for dysplasia.³⁹

Current endoscopic resection techniques

Endoscopic resection techniques include mucosal resection, submucosal dissection, radio­frequency ablation, cryotherapy, argon plasma coagulation, and photodynamic therapy (Figure 2).

In mucosal resection, the lesion is either suctioned into a band ligator, after which a band is placed around the lesion, or suctioned into a cap fitted at the end of the endoscope, after which the lesion is removed using a snare.

In submucosal dissection, a liquid is injected into the submucosa to lift the lesion, making it easier to remove. The procedure is technically complex and requires additional training.

In radiofrequency ablation, a special catheter is passed through the endoscope to ablate the affected epithelium by thermal injury. Argon plasma coagulation works in a similar way, but uses ionized argon gas to induce thermal coagulation of metaplastic epithelium.

Cryotherapy produces cellular injury by rapid freezing and thawing of tissue using a cryogen such as liquid nitrogen or nitrous oxide.

In photodynamic therapy, a photosensitizer (porfimer sodium) is administered and taken up preferentially by metaplastic epithelium; it is then activated by transmission of red light using the endoscope, leading to destruction of the metaplastic epithelium.

Of the different techniques, radiofrequency ablation has the most evidence for efficacy and hence is the most commonly used.

All of these procedures are generally well tolerated and have favorable side-effect profiles. After radiofrequency ablation with or without mucosal resection, esophageal strictures are noted in 5.6% of patients, and bleeding and perforation occur rarely (1% and 0.6% of patients, respectively).⁴⁰ Submucosal dissection is associated with a higher rate of complications such as stricture formation (11% of patients) and bleeding or perforation (1.5% of patients).⁴¹

Most patients with low-grade dysplasia (73%) are down-staged to nondysplastic Barrett esophagus or to indefinite for dysplasia after review by expert pathologists.⁴² Patients with confirmed and persistent low-grade dysplasia are at higher risk of progression.⁴³

Once low-grade dysplasia is confirmed by a second gastrointestinal pathologist, the patient should undergo endoscopic ablation. A landmark study by Shaheen et al⁴⁴ demonstrated the benefit of radiofrequency ablation in achieving complete eradication of dysplasia (90.5% vs 22.7% for a sham procedure) and complete eradication of intestinal metaplasia (77.4% vs 2.3% for a sham procedure). In another trial of 136 patients with low-grade dysplasia followed for 3 years, Phoa et al⁴⁵ demonstrated that radiofrequency ablation reduced the rate of progression to high-grade dysplasia by 25% and to adenocarcinoma by 7.4% compared with endoscopic surveillance.

Patients with confirmed low-grade dysplasia who do not undergo eradication therapy should have surveillance endoscopy every 6 to 12 months (Table 1).

HIGH-GRADE DYSPLASIA: RECOMMENDED MANAGEMENT

As with low-grade dysplasia, the diagnosis of high-grade dysplasia needs to be confirmed by a second pathologist with gastrointestinal expertise. In the past, the treatment was esophagectomy, but due to lower morbidity and equivalent efficacy of radiofrequency ablation,⁴⁶ the current treatment of choice is endoscopic mucosal resection of raised lesions, followed by radiofrequency ablation of the entire affected segment.

In the study by Shaheen et al,⁴⁴ 42 patients with high-grade dysplasia were randomized to radiofrequency ablation and 21 to a sham procedure, and 81% of ablation patients achieved complete eradication of dysplasia vs 19% with the sham procedure. Eradication of intestinal metaplasia was achieved in 77% of ablation patients vs 2% of patients with the sham therapy. Results of 3-year follow-up from the same cohort showed complete eradication of dysplasia in 98% and of intestinal metaplasia in 91%.⁴⁷

Endoscopic eradication therapy is recommended for all patients with Barrett esophagus and high-grade dysplasia without a life-limiting comorbidity. Alternatively, surveillance every 3 months is an option if the patient does not wish to undergo eradication therapy. Radiofrequency ablation is more cost-effective than esophagectomy or endoscopic surveillance followed by treatment once patients develop adenocarcinoma.^48,49

EARLY ESOPHAGEAL ADENOCARCINOMA: RECOMMENDED MANAGEMENT

Adenocarcinoma limited to the mucosa and without evidence of nodal involvement can be resected endoscopically. In patients with localized cancer, mucosal resection is done not only for therapeutic purposes but also for staging. Ideal management is multidisciplinary, including a gastroenterologist, thoracic surgeon, oncologist, pathologist, and radiation oncologist.

If lesions have features suggesting submucosal invasion or are greater than 1.5 cm in size, or if it is difficult to separate (ie, lift) the mucosa from the submucosal layer with injection of saline, then submucosal dissection is recommended.⁵⁰ Because of the risk of metachronous lesions, ablation of the remaining Barrett esophagus mucosa is recommended after resection of cancer.

Endoscopic eradication is highly effective and durable for the treatment of intramucosal esophageal adenocarcinoma. In a study of 1,000 patients, 963 patients (96.3%) had achieved a complete response; 12 patients (3.7%) underwent surgery after eradication failed during a follow-up of almost 5 years.⁵¹ Metachronous lesions or recurrence of cancer developed during the follow-up period in 140 patients (14.5%) but were successfully treated endoscopically in 115, resulting in a long-term complete remission rate of 93.8%.

POSTABLATION MANAGEMENT

Because of the risk of recurrence of dysplasia after ablation, long-term PPI therapy and surveillance are recommended.

Surveillance endoscopy involves 4-quadrant biopsies taken every 1 cm from the entire length of segment where Barrett esophagus had been seen before ablation.

The timing of surveillance intervals depends on the preablation grade of dysplasia. For low-grade dysplasia, the recommendation is every 6 months for the first year after ablation and, if there is no recurrence of dysplasia, annually after that.² After treatment of high-grade dysplasia or intramucosal adenocarcinoma, the recommendation is every 3 months for the first year, every 6 months in the second year, and then annually.²

All cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus.¹ But most cases of Barrett esophagus go undiagnosed. And Barrett esophagus is seen in 5% to 15% of patients with gastroesophageal reflux disease.² These facts clearly emphasize the need for screening. Here, we review the rationale and recommendations for screening and surveillance, as well as the range of treatment options.

SCOPE OF THE PROBLEM

The American Cancer Society estimated there were 17,290 new cases of esophageal cancer and 15,850 deaths from it in the United States in 2018.³ Of the 2 main histologic types of esophageal cancer, adenocarcinoma and squamous cell cancer, adenocarcinoma is more common in the United States.

The precursor lesion is Barrett esophagus, defined as an extension of salmon-colored mucosa at least 1 cm into the tubular esophagus proximal to the gastroesophageal junction, with biopsy confirmation of intestinal metaplasia.⁴

The natural course of progression to dysplasia and cancer in Barrett esophagus is unknown but is thought to be stepwise, from no dysplasia to low-grade dysplasia to high-grade dysplasia and cancer, and the cancer risk depends on the degree of dysplasia: the annual risk is 0.33% if there is no dysplasia, 0.54% with low-grade dysplasia, and 7% with high-grade dysplasia.⁴

Although all cases of esophageal adenocarcinoma are thought to arise from Barrett esophagus,¹ more than 90% of patients with newly diagnosed esophageal adenocarcinoma do not have a prior diagnosis of Barrett esophagus.⁵ Therefore, there is a substantial unmet need to expand screening for Barrett esophagus in people at risk.

GASTROESOPHAGEAL REFLUX DISEASE IS A RISK FACTOR FOR CANCER

The rationale behind screening is that detecting Barrett esophagus early and intervening in a timely manner in patients at higher risk of developing adenocarcinoma will decrease mortality.

Chronic gastroesophageal reflux disease is a strong risk factor for esophageal adenocarcinoma (odds ratio [OR] 7.7, 95% confidence interval [CI] 5.3–11.4), and the risk increases when symptoms are long-standing (> 20 years) or severe (OR 43.5, 95% CI 18.3–103.5) or occur daily (OR 5.5, 95% CI 3.2–9.3).⁶

Reflux symptoms are scored as follows:

• Heartburn only, 1 point
• Regurgitation only, 1 point
• Heartburn with regurgitation, 1.5 points
• Nightly symptoms (2 points if yes, 0 if no)
• Symptoms once a week, 0 points; 2 to 6 times a week, 1 point; 7 to 15 times a week, 2 points; more than 15 times a week, 3 points.⁶

A score of 4.5 or higher indicates severe reflux disease. However, it is worth noting that the annual incidence of esophageal adenocarcinoma in patients with long-term gastroesophageal reflux disease is less than 0.001%.⁷

RISK FACTORS FOR BARRETT ESOPHAGUS

Risk factors for Barrett esophagus include:

Male sex. Barrett esophagus is more prevalent in men than in women, at a ratio of 2 to 1; but in individuals under age 50, the ratio is 4 to 1.⁸

Age 50 or older. Barrett esophagus is usually diagnosed in the sixth to seventh decade of life, and the prevalence increases from 2.1% in the third decade to 9.3% in the sixth decade.⁹

White race. It is more prevalent in whites than in blacks (5.0% vs 1.5%, P < .0001).¹⁰

Central obesity. Waist circumference is an independent risk factor: every 5-cm increase carries an OR of 1.14 (95% CI 1.03–1.27, P = .02).¹¹

Cigarette smoking increases the risk of Barrett esophagus (OR 1.42; 95% CI 1.15–1.76).¹²

A family history of Barrett esophagus or esophageal adenocarcinoma is a strong risk factor (OR 12, 95% CI 3.3–44.8). In 1 study, the risk in first- and second-degree relatives of patients with Barrett esophagus was 24%, compared with 5% in a control population (P < .005).¹³

SCREENING GUIDELINES AND DRAWBACKS

The standard screening method is esophagogastroduodenoscopy with sedation, with careful visual inspection and 4-quadrant biopsies every 2 cm using the Seattle protocol, ie, including biopsy of any mucosal irregularities in salmon-colored mucosa above the gastroesophageal junction (Figure 1).⁴

Endoscopic screening is cost-effective, costing $10,440 per quality-adjusted life-year saved, which is well below the accepted threshold of less than $100,000.¹⁴ However, it is still expensive, invasive, and not ideal for screening large populations.

Less-invasive methods under study

Less-invasive, less-expensive methods being tested for mass screening include:

Unsedated transnasal endoscopy. Done with only topical anesthesia, it has high diagnostic accuracy and is quicker and more cost-effective than standard esophagogastroduodenoscopy, with fewer adverse effects. However, the procedure has not yet gained widespread acceptance for regular use by gastroenterologists.¹⁵

A swallowable sponge. Another promising test is cell collection using the Cytosponge Cell Collection Device (Medtronic, Minneapolis, MN). An encapsulated compressed sponge with a string attached is swallowed; in the stomach, the capsule dissolves, and the sponge expands and is then withdrawn using the attached string. The obtained cytology sample from the lower esophagus is then tested for trefoil factor 3, a protein biomarker for Barrett esophagus.¹⁶

A retractable balloon. The EsoCheck Cell Collection Device is a retractable balloon attached to a string. When swallowed, it gathers distal esophageal cells for detecting methylated DNA markers for Barrett esophagus.¹⁷

Esophageal capsule endoscopy uses a camera to visualize the esophagus, but lacks the ability to obtain biopsy samples.

Other screening methods are being tested, although data are limited. Liquid biopsy uses a blood sample to detect microRNAs that are dysregulated in cancer. The “electronic nose” is a device that detects exhaled volatile organic compounds altered in Barrett esophagus. Another test involves taking an oral wash sample to study the oral microbiome for a pattern associated with adenocarcinoma.^18–21

Surveillance in Barrett esophagus aims to detect premalignant changes or early-stage adenocarcinoma to provide longer survival and lower cancer-related mortality. Recent evidence suggests that patients with esophageal adenocarcinoma that is diagnosed in a Barrett esophagus surveillance program have an earlier stage of disease and therefore a survival benefit.²²

Patient education is essential

Before enrolling a patient in a surveillance program, the clinician should explain the risks, benefits, and limitations, the importance of periodic endoscopy, and the possible eventual need for endoscopic therapy or surgery.

The endoscopic procedure

Surveillance involves examination by high-definition white-light endoscopy, with random 4-quadrant biopsies every 2 cm (or every 1 cm in patients with a history of dysplasia) and biopsy of any mucosal irregularity (nodule, ulcer, or other visible lesion). The degree of dysplasia determines the frequency of follow-up surveillance intervals and the need for endoscopic eradication therapy, as presented in professional society guidelines (Table 1).^4,23,24

Advanced methods for detecting dysplasia

Newer endoscopic surveillance techniques include dye-based chromoendoscopy, narrow-band imaging, confocal laser endomicroscopy, volumetric laser endomicroscopy, and wide-area transepithelial sampling with computer-assisted 3-dimensional analysis. All these techniques are used to increase the detection of dysplasia. Of these, dye-based chromoendoscopy, narrow-band imaging, and confocal laser endomicroscopy meet current criteria of the American Society for Gastrointestinal Endoscopy for preservation and incorporation of valuable endoscopic innovations.²³

MANAGEMENT OF NONDYSPLASTIC BARRETT ESOPHAGUS

A proton pump inhibitor (PPI) is recommended to control reflux symptoms in patients with nondysplastic Barrett esophagus. But it is important to counsel patients on additional ways to protect against esophageal adenocarcinoma, such as:

• Low to moderate alcohol consumption
• Regular physical activity
• Increased dietary intake of fruits, vegetables, folate, fiber, beta-carotene, and vitamin C
• Weight control
• Smoking cessation.²⁵

Surveillance endoscopy with 4-quadrant biopsies at 2-cm intervals is recommended every 3 to 5 years (Table 1).

DOES CHEMOPREVENTION HAVE A ROLE?

Chemoprevention is an exciting area of research in preventing progression to adenocarcinoma in patients with Barrett esophagus. Various drugs such as aspirin, other nonsteroidal anti-inflammatory drugs (NSAIDs), PPIs, metformin, and statins have been studied.

Aspirin

Aspirin has been shown to prevent development of Barrett esophagus in patients with reflux disease,²⁶ but more studies are needed to validate those findings.

PPIs

Gastroesophageal reflux disease is a primary risk factor for esophageal adenocarcinoma, and gastric acid suppression with PPIs reduces cancer risk. PPI therapy is associated with a 71% decrease in the risk of high-grade dysplasia and adenocarcinoma in patients with Barrett esophagus (OR 0.29, 95% CI 0.12–0.79).²⁷ Long-term therapy (> 2 to 3 years) has a higher protective effect (adjusted OR 0.45, 95% CI 0.19–1.06) than short-term therapy (< 2 to 3 years) (adjusted OR 1.09, 95% CI 0.47–2.56).²⁷

NSAIDs

NSAIDs, including aspirin, have been associated with decreased risk of colon, stomach, lung, breast, and esophageal cancer due to their potential to inhibit cyclooxygenase 2 (COX-2) enzymes.

A meta-analysis demonstrated that aspirin and NSAIDs led to a 32% reduction in the risk of adenocarcinoma (OR 0.68, 95% CI 0.56–0.83). The benefit was even greater if the drug was taken daily or more frequently (OR 0.56, 95% CI 0.43–0.73, P < .001) or was taken for 10 or more years (OR 0.63, 95% CI 0.45–0.90, P = .04).²⁸

PPI plus aspirin

The best evidence for the role of PPIs and aspirin in reducing the risk of dysplasia comes from the Aspirin and Esomeprazole Chemoprevention in Barrett’s Metaplasia Trial.²⁹ This randomized, controlled trial compared 4 regimens consisting of esomeprazole (a PPI) in either a high dose (40 mg twice daily) or a low dose (20 mg once daily) plus either aspirin (300 or 320 mg per day) or no aspirin in 2,557 patients with Barrett esophagus. The composite end point was the time to all-cause mortality, adenocarcinoma, or high-grade dysplasia.

At a median follow-up of 8.9 years, the combination of high-dose esomeprazole plus aspirin had the strongest effect compared with low-dose esomeprazole without aspirin (time ratio 1.59, 95% CI 1.14–2.23, P = .0068). The number needed to treat was 34 for esomeprazole and 43 for aspirin.²⁹

Based on these data, we can conclude that aspirin and PPIs can prevent dysplasia and all-cause mortality in Barrett esophagus.

Metformin: No evidence of benefit

Metformin was studied as a protective agent against obesity-associated cancers including esophageal adenocarcinoma, as it reduces insulin levels.

In a randomized controlled trial³⁰ in 74 patients with Barrett esophagus, metformin (starting at 500 mg daily, increasing to 2,000 mg/day by week 4) was compared with placebo. At 12 weeks, the percent change in esophageal levels of the biomarker pS6K1—an intracellular mediator of insulin and insulin-like growth factor activation in Barrett epithelium—did not differ significantly between the 2 groups (1.4% with metformin vs −14.7% with placebo; 1-sided P = .80). This suggested that metformin did not significantly alter proliferation or apoptosis in Barrett epithelium, despite reducing serum insulin levels and insulin resistance. Thus, metformin did not demonstrate a chemoprotective effect in preventing the progression of Barrett esophagus to adenocarcinoma.

Vitamin D affects genes regulating proliferation, apoptosis, and differentiation, and has therefore been studied as a potential antineoplastic agent. Its deficiency has also been associated with increased risk of esophageal adenocarcinoma. However, its efficacy in chemoprevention is unclear.³¹

One study found no association between serum 25-hydroxyvitamin D levels and prevalence of dysplasia in Barrett esophagus (P = .90). An increase in vitamin D levels had no effect on progression to dysplasia or cancer (for every 5-nmol/L increase from baseline, hazard ratio 0.98, P = .62).³²

In another study, supplementation with vitamin D₃ (cholecalciferol 50,000 IU weekly) plus a PPI for 12 weeks significantly improved the serum 25-hydroxyvitamin D levels without significant changes in gene expression from Barrett epithelium.³³ These findings were confirmed in a meta-analysis that showed no consistent association between vitamin D exposure and risk of esophageal neoplasm.³⁴

Thus, there is currently no evidence to support vitamin D for chemoprevention in Barrett esophagus or esophageal adenocarcinoma.

Statins

In addition to lowering cholesterol, statins have antiproliferative, pro-apoptotic, anti-angiogenic, and immunomodulatory effects that prevent cancer, leading to a 41% reduction in the risk of adenocarcinoma in patients with Barrett esophagus in one study (adjusted OR 0.59, 95% CI 0.45–0.78); the number needed to treat with statins to prevent 1 case of adenocarcinoma was 389.³⁵

A meta-analysis also showed that statin use was associated with a lower risk of progression of Barrett esophagus (OR 0.48, 95% CI 0.31–0.73).³⁶

In general, statins appear promising for chemoprevention, but more study is needed.

When is chemoprevention appropriate?

Chemoprevention is not recommended for all patients with Barrett esophagus, given that the condition affects 1% to 2% of the US adult population, and very few patients have progression to esophageal adenocarcinoma. Rather, chemoprevention may be considered in patients with Barrett esophagus and multiple risk factors for adenocarcinoma.

INDEFINITE DYSPLASIA

In Barrett esophagus with indefinite dysplasia, either the epithelial abnormalities are insufficient for a diagnosis of dysplasia, or the nature of the epithelial abnormalities is uncertain due to inflammation or technical difficulties with specimen processing. The risk of high-grade dysplasia or cancer within 1 year of the diagnosis of indefinite dysplasia varies between 1.9% and 15%.³⁷ The recommendation for management is to optimize acid-suppressive therapy for 3 to 6 months and then to repeat esophagogastroduodenoscopy. If indefinite dysplasia is noted again, repeat endoscopy in 12 months is recommended.²

ENDOSCOPIC ERADICATION: AN OVERVIEW

Because dysplasia in Barrett esophagus carries a high risk of progression to cancer, the standard of care is endoscopic mucosal resection of visible lesions, followed by ablation of the flat mucosa, with the aim of achieving complete eradication of intestinal metaplasia.^4,38 The initial endoscopic treatment is followed by outpatient sessions every 8 to 10 weeks until the dysplasia is eradicated. A key part of treatment during this time is maximal acid suppression with a PPI twice daily and a histamine-2 blocker at night. In rare cases, fundoplication is required to control reflux refractory to medical therapy.

After eradication is confirmed, continued surveillance is necessary, as recurrences have been reported at a rate of 4.8% per year for intestinal metaplasia, and 2% per year for dysplasia.³⁹

Current endoscopic resection techniques

Endoscopic resection techniques include mucosal resection, submucosal dissection, radio­frequency ablation, cryotherapy, argon plasma coagulation, and photodynamic therapy (Figure 2).

In mucosal resection, the lesion is either suctioned into a band ligator, after which a band is placed around the lesion, or suctioned into a cap fitted at the end of the endoscope, after which the lesion is removed using a snare.

In submucosal dissection, a liquid is injected into the submucosa to lift the lesion, making it easier to remove. The procedure is technically complex and requires additional training.

In radiofrequency ablation, a special catheter is passed through the endoscope to ablate the affected epithelium by thermal injury. Argon plasma coagulation works in a similar way, but uses ionized argon gas to induce thermal coagulation of metaplastic epithelium.

Cryotherapy produces cellular injury by rapid freezing and thawing of tissue using a cryogen such as liquid nitrogen or nitrous oxide.

In photodynamic therapy, a photosensitizer (porfimer sodium) is administered and taken up preferentially by metaplastic epithelium; it is then activated by transmission of red light using the endoscope, leading to destruction of the metaplastic epithelium.

Of the different techniques, radiofrequency ablation has the most evidence for efficacy and hence is the most commonly used.

All of these procedures are generally well tolerated and have favorable side-effect profiles. After radiofrequency ablation with or without mucosal resection, esophageal strictures are noted in 5.6% of patients, and bleeding and perforation occur rarely (1% and 0.6% of patients, respectively).⁴⁰ Submucosal dissection is associated with a higher rate of complications such as stricture formation (11% of patients) and bleeding or perforation (1.5% of patients).⁴¹

Most patients with low-grade dysplasia (73%) are down-staged to nondysplastic Barrett esophagus or to indefinite for dysplasia after review by expert pathologists.⁴² Patients with confirmed and persistent low-grade dysplasia are at higher risk of progression.⁴³

Once low-grade dysplasia is confirmed by a second gastrointestinal pathologist, the patient should undergo endoscopic ablation. A landmark study by Shaheen et al⁴⁴ demonstrated the benefit of radiofrequency ablation in achieving complete eradication of dysplasia (90.5% vs 22.7% for a sham procedure) and complete eradication of intestinal metaplasia (77.4% vs 2.3% for a sham procedure). In another trial of 136 patients with low-grade dysplasia followed for 3 years, Phoa et al⁴⁵ demonstrated that radiofrequency ablation reduced the rate of progression to high-grade dysplasia by 25% and to adenocarcinoma by 7.4% compared with endoscopic surveillance.

Patients with confirmed low-grade dysplasia who do not undergo eradication therapy should have surveillance endoscopy every 6 to 12 months (Table 1).

HIGH-GRADE DYSPLASIA: RECOMMENDED MANAGEMENT

As with low-grade dysplasia, the diagnosis of high-grade dysplasia needs to be confirmed by a second pathologist with gastrointestinal expertise. In the past, the treatment was esophagectomy, but due to lower morbidity and equivalent efficacy of radiofrequency ablation,⁴⁶ the current treatment of choice is endoscopic mucosal resection of raised lesions, followed by radiofrequency ablation of the entire affected segment.

In the study by Shaheen et al,⁴⁴ 42 patients with high-grade dysplasia were randomized to radiofrequency ablation and 21 to a sham procedure, and 81% of ablation patients achieved complete eradication of dysplasia vs 19% with the sham procedure. Eradication of intestinal metaplasia was achieved in 77% of ablation patients vs 2% of patients with the sham therapy. Results of 3-year follow-up from the same cohort showed complete eradication of dysplasia in 98% and of intestinal metaplasia in 91%.⁴⁷

Endoscopic eradication therapy is recommended for all patients with Barrett esophagus and high-grade dysplasia without a life-limiting comorbidity. Alternatively, surveillance every 3 months is an option if the patient does not wish to undergo eradication therapy. Radiofrequency ablation is more cost-effective than esophagectomy or endoscopic surveillance followed by treatment once patients develop adenocarcinoma.^48,49

EARLY ESOPHAGEAL ADENOCARCINOMA: RECOMMENDED MANAGEMENT

Adenocarcinoma limited to the mucosa and without evidence of nodal involvement can be resected endoscopically. In patients with localized cancer, mucosal resection is done not only for therapeutic purposes but also for staging. Ideal management is multidisciplinary, including a gastroenterologist, thoracic surgeon, oncologist, pathologist, and radiation oncologist.

If lesions have features suggesting submucosal invasion or are greater than 1.5 cm in size, or if it is difficult to separate (ie, lift) the mucosa from the submucosal layer with injection of saline, then submucosal dissection is recommended.⁵⁰ Because of the risk of metachronous lesions, ablation of the remaining Barrett esophagus mucosa is recommended after resection of cancer.

Endoscopic eradication is highly effective and durable for the treatment of intramucosal esophageal adenocarcinoma. In a study of 1,000 patients, 963 patients (96.3%) had achieved a complete response; 12 patients (3.7%) underwent surgery after eradication failed during a follow-up of almost 5 years.⁵¹ Metachronous lesions or recurrence of cancer developed during the follow-up period in 140 patients (14.5%) but were successfully treated endoscopically in 115, resulting in a long-term complete remission rate of 93.8%.

POSTABLATION MANAGEMENT

Because of the risk of recurrence of dysplasia after ablation, long-term PPI therapy and surveillance are recommended.

Surveillance endoscopy involves 4-quadrant biopsies taken every 1 cm from the entire length of segment where Barrett esophagus had been seen before ablation.

The timing of surveillance intervals depends on the preablation grade of dysplasia. For low-grade dysplasia, the recommendation is every 6 months for the first year after ablation and, if there is no recurrence of dysplasia, annually after that.² After treatment of high-grade dysplasia or intramucosal adenocarcinoma, the recommendation is every 3 months for the first year, every 6 months in the second year, and then annually.²

Migraine is the second leading cause of years of life lived with a disability globally.¹ It affects people of all ages, but particularly during the years associated with the highest productivity in terms of work and family life.

Migraine is a genetic neurologic disease that can be influenced or triggered by environmental factors. However, triggers do not cause migraine. For example, stress does not cause migraine, but it can exacerbate it.

Primary care physicians can help patients reduce the likelihood of a migraine attack, the severity of symptoms, or both by offering lifestyle counseling centered around the mnemonic SEEDS: sleep, exercise, eat, diary, and stress. In this article, each factor is discussed individually for its current support in the literature along with best-practice recommendations.

S IS FOR SLEEP

Before optimizing sleep hygiene, screen for sleep apnea, especially in those who have chronic daily headache upon awakening. An excellent tool is the STOP-Bang screening questionnaire⁵ (www.stopbang.ca/osa/screening.php). Patients respond “yes” or “no” to the following questions:

• Snoring: Do you snore loudly (louder than talking or loud enough to be heard through closed doors)?
• Tired: Do you often feel tired, fatigued, or sleepy during the daytime?
• Observed: Has anyone observed you stop breathing during your sleep?
• Pressure: Do you have or are you being treated for high blood pressure?
• Body mass index greater than 35 kg/m²?
• Age over 50?
• Neck circumference larger than 40 cm (females) or  42 cm (males)?
• Gender—male?

Each “yes” answer is scored as 1 point. A score less than 3 indicates low risk of obstructive sleep apnea; 3 to 4 indicates moderate risk; and 5 or more indicates high risk. Optimization of sleep apnea with continuous positive airway pressure therapy can improve sleep apnea headache.⁶ The improved sleep from reduced arousals may also mitigate migraine symptoms.

Behavioral modification for sleep hygiene can convert chronic migraine to episodic migraine.⁷ One such program is stimulus control therapy, which focuses on using cues to initiate sleep (Table 1). Patients are encouraged to keep the bedroom quiet, dark, and cool, and to go to sleep at the same time every night. Importantly, the bed should be associated only with sleep. If patients are unable to fall asleep within 20 to 30 minutes, they should leave the room so they do not associate the bed with frustration and anxiety. Use of phones, tablets, and television in the bedroom is discouraged as these devices may make it more difficult to fall asleep.⁸

The next option is sleep restriction, which is useful for comorbid insomnia. Patients keep a sleep diary to better understand their sleep-wake cycle. The goal is 90% sleep efficiency, meaning that 90% of the time in bed (TIB) is spent asleep. For example, if the patient is in bed 8 hours but asleep only 4 hours, sleep efficiency is 50%. The goal is to reduce TIB to match the time asleep and to agree on a prescribed daily wake-up time. When the patient is consistently sleeping 90% of the TIB, add 30-minute increments until he or she is appropriately sleeping 7 to 8 hours at night.⁹ Naps are not recommended.

Let patients know that their migraine may worsen until a new routine sleep pattern emerges. This method is not recommended for patients with untreated sleep apnea.

E IS FOR EXERCISE

Exercise is broadly recommended for a healthy lifestyle; some evidence suggests that it can also be useful in the management of migraine.¹⁰ Low levels of physical activity and a sedentary lifestyle are associated with migraine.¹¹ It is unclear if patients with migraine are less likely to exercise because they want to avoid triggering a migraine or if a sedentary lifestyle increases their risk.

Exercise has been studied for its prophylactic benefits in migraine, and one hypothesis relates to beta-endorphins. Levels of beta-endorphins are reduced in the cerebrospinal fluid of patients with migraine.¹² Exercise programs may increase levels while reducing headache frequency and duration.¹³ One study showed that pain thresholds do not change with exercise programs, suggesting that it is avoidance behavior that is positively altered rather than the underlying pain pathways.¹⁴

A systematic review and meta-analysis based on 5 randomized controlled trials and 1 nonrandomized controlled clinical trial showed that exercise reduced monthly migraine days by only 0.6 (± 0.3) days, but the data also suggested that as the exercise intensity increased, so did the positive effects.¹⁰

Some data suggest that exercise may also reduce migraine duration and severity as well as the need for abortive medication.¹⁰ Two studies in this systematic review^15,16 showed that exercise benefits were equivalent to those of migraine preventives such as amitriptyline and topiramate; the combination of amitriptyline and exercise was more beneficial than exercise alone. Multiple types of exercise were beneficial, including walking, jogging, cross-training, and cycling when done for least 6 weeks and for 30 to 50 minutes 3 to 5 times a week.

These findings are in line with the current recommendations for general health from the American College of Sports Medicine, ie, moderate to vigorous cardio­respiratory exercise for 30 to 60 minutes 3 to 5 times a week (or 150 minutes per week). The daily exercise can be continuous or done in intervals of less than 20 minutes. For those with a sedentary lifestyle, as is seen in a significant proportion of the migraine population, light to moderate exercise for less than 20 minutes is still beneficial.¹⁷

Based on this evidence, the best current recommendation for patients with migraine is to engage in graded moderate cardiorespiratory exercise, although any exercise is better than none. If a patient is sedentary or has poor exercise tolerance, or both, exercising once a week for shorter time periods may be a manageable place to start.

Some patients may identify exercise as a trigger or exacerbating factor in migraine. These patients may need appropriate prophylactic and abortive therapies before starting an exercise regimen.

THE SECOND E IS FOR EAT (FOOD AND DRINK)

Many patients believe that some foods trigger migraine attacks, but further study is needed. The most consistent food triggers appear to be red wine and caffeine (withdrawal).^18,19 Interestingly, patients with migraine report low levels of alcohol consumption,²⁰ but it is unclear if that is because alcohol has a protective effect or if patients avoid it.

Some patients may crave certain foods in the prodromal phase of an attack, eat the food, experience the attack, and falsely conclude that the food caused the attack.²¹ Premonitory symptoms include fatigue, cognitive changes, homeostatic changes, sensory hyperresponsiveness, and food cravings.²¹ It is difficult to distinguish between premonitory phase food cravings and true triggers because premonitory symptoms can precede headache by 48 to 72 hours, and the timing for a trigger to be considered causal is not known.²²

Chocolate is often thought to be a migraine trigger, but the evidence argues against this and even suggests that sweet cravings are a part of the premonitory phase.²³ Monosodium glutamate is often identified as a trigger as well, but the literature is inconsistent and does not support a causal relationship.²⁴ Identifying true food triggers in migraine is difficult, and patients with migraine may have poor quality diets, with some foods acting as true triggers for certain patients.²⁵ These possibilities have led to the development of many “migraine diets,” including elimination diets.

Elimination diets

Elimination diets involve avoiding specific food items over a period of time and then adding them back in one at a time to gauge whether they cause a reaction in the body. A number of these diets have been studied for their effects on headache and migraine:

Gluten-free diets restrict foods that contain wheat, rye, and barley. A systematic review of gluten-free diets in patients with celiac disease found that headache or migraine frequency decreased by 51.6% to 100% based on multiple cohort studies (N = 42,388).²⁶ There are no studies on the use of a gluten-free diet for migraine in patients without celiac disease.

Immunoglobulin G-elimination diets restrict foods that serve as antigens for IgG. However, data supporting these diets are inconsistent. Two small randomized controlled trials found that the diets improved migraine symptoms, but a larger study found no improvement in the number of migraine days at 12 weeks, although there was an initially significant effect at 4 weeks.^27–29

Antihistamine diets restrict foods that have high levels of histamines, including fermented dairy, vegetables, soy products,  wine, beer, alcohol, and those that cause histamine release regardless of IgE testing results. A prospective single-arm study of antihistamine diets in patients with chronic headache reported symptom improvement, which could be applied to certain comorbidities such as  mast cell activation syndrome.³⁰ Another prospective nonrandomized controlled study eliminated foods based on positive IgE skin-prick testing for allergy in patients with recurrent migraine and found that it reduced headache frequency.³¹

Tyramine-free diets are often recommended due to the presumption that tyramine-containing foods (eg, aged cheese, cured or smoked meats and fish, and beer) are triggers. However, multiple studies have reviewed this theory with inconsistent results,³² and the only study of a tyramine-free diet was negative.³³ In addition, commonly purported high-tyramine foods have lower tyramine levels than previously thought.³⁴

Low-fat diets in migraine are supported by 2 small randomized controlled trials and a prospective study showing a decrease in symptom severity; the results for frequency are inconsistent.^35–37

Low-glycemic index diets are supported in migraine by 1 randomized controlled trial that showed improvement in migraine frequency in a diet group and in a control group of patients who took a standard migraine-preventive medication to manage their symptoms.³⁸

Other migraine diets

Diets high in certain foods or ingredient ratios, as opposed to elimination diets, have also been studied in patients with migraine. One promising diet containing high levels of omega-3 fatty acids and low levels of omega-6 fatty acids was shown in a systematic review to reduce the duration of migraine but not the frequency or severity.³⁹ A more recent randomized controlled trial of this diet in chronic migraine also showed that it decreased migraine frequency.⁴⁰

The ketogenic diet (high fat, low carbohydrate) had promising results in a randomized controlled trial in overweight women with migraine and in a prospective study.^41,42 However, a prospective study of the Atkins diet in teenagers with chronic daily headaches showed no benefit.⁴³ The ketogenic diet is difficult to follow and may work in part due to weight loss alone, although ketogenesis itself may also play a role.^41,44

Sodium levels have been shown to be higher in the cerebrospinal fluid of patients with migraine than in controls, particularly during an attack.⁴⁵ For a prehypertensive population or an elderly population, a low-sodium diet may be beneficial based on 2 prospective trials.^46,47 However, a younger female population without hypertension and low-to-normal body mass index had a reduced probability of migraine while consuming a high-sodium diet.⁴⁸

Counseling about sodium intake should be tailored to specific patient populations. For example, a diet low in sodium may be appropriate for patients with vascular risk factors such as hypertension, whereas a high-sodium diet may be appropriate in patients with comorbidities like postural tachycardia syndrome or in those with a propensity for low blood pressure or low body mass index.

Encourage routine meals and hydration

The standard advice for patients with migraine is to consume regular meals. Headaches have been associated with fasting, and those with migraine are predisposed to attacks in the setting of fasting.^49,50 Migraine is more common when meals are skipped, particularly breakfast.⁵¹

It is unclear how fasting lowers the migraine threshold. Nutritional studies show that skipping meals, particularly breakfast, increases low-grade inflammation and impairs  glucose metabolism by affecting insulin and fat oxidation metabolism.⁵² However, hypoglycemia itself is not a consistent cause of headache or migraine attacks.⁵³ As described above, a randomized controlled trial of a low-glycemic index diet actually decreased migraine frequency and severity.³⁸ Skipping meals also reduces energy and is associated with reduced physical activity, perhaps leading to multiple compounding triggers that further lower the migraine threshold.^54,55

When counseling patients about the need to eat breakfast, consider what they normally consume (eg, is breakfast just a cup of coffee?). Replacing simple carbohydrates with protein, fats, and fiber may be beneficial for general health, but the effects on migraine are not known, nor is the optimal composition of breakfast foods.⁵⁵

The optimal timing of breakfast relative to awakening is also unclear, but in general, it should be eaten within 30 to 60 minutes of rising. Also consider patients’ work hours—delayed-phase or shift workers have altered sleep cycles.

Recommendations vary in regard to hydration. Headache is associated with fluid restriction and dehydration,^56,57 but only a few studies suggest that rehydration and increased hydration status can improve migraine.⁵⁸ In fact, a single post hoc analysis of a metoclopramide study showed that intravenous fluid alone for patients with migraine in the emergency room did not improve pain outcomes.⁵⁹

The amount of water patients should drink daily in the setting of migraine is also unknown, but a study showed benefit with 4 L, which equates to a daily intake of 16 eight-ounce glasses.⁶⁰ One review on general health that could be extrapolated given the low risk of the intervention indicated that 1.8 L daily (7 to 8 eight-ounce glasses) promoted a euhydration status in most people, although many factors contribute to hydration status.⁶¹

Caffeine intake is also a major consideration. Caffeine is a nonspecific adenosine receptor antagonist that modulates adenosine receptors like the pronociceptive 2A receptor, leading to changes integral to the neuropathophysiology of migraine.⁶² Caffeine has analgesic properties at doses greater than 65 to 200 mg and augments the effects of analgesics such as acetaminophen and aspirin. Chronic caffeine use can lead to withdrawal symptoms when intake is stopped abruptly; this is thought to be due to upregulation of adenosine receptors, but the effect varies based on genetic predisposition.¹⁹

The risk of chronic daily headache may relate to high use of caffeine preceding the onset of chronification, and caffeine abstinence may improve response to acute migraine treatment.^19,63 There is a dose-dependent risk of headache.^64,65 Current recommendations suggest limiting caffeine consumption to less than 200 mg per day or stopping caffeine consumption altogether based on the quantity required for caffeine-withdrawal headache.⁶⁶ Varying  the caffeine dose from day to day may also trigger headache due to the high sensitivity to caffeine withdrawal.

While many diets have shown potential benefit in patients with migraine, more studies are needed before any one “migraine diet” can be recommended. Caution should be taken, as there is risk of adverse effects from nutrient deficiencies or excess levels, especially if the patient is not under the care of a healthcare professional who is familiar with the diet.

Whether it is beneficial to avoid specific food triggers at this time is unclear and still controversial even within the migraine community because some of these foods may be misattributed as triggers instead of premonitory cravings driven by the hypothalamus. It is important to counsel patients with migraine to eat a healthy diet with consistent meals, to maintain adequate hydration, and to keep their caffeine intake low or at least consistent, although these teachings are predominantly based on limited studies with extrapolation from nutrition research.

D IS FOR DIARY

A headache diary is a recommended part of headache management and may enhance the accuracy of diagnosis and assist in treatment modifications. Paper and electronic diaries have been used. Electronic diaries may be more accurate for real-time use, but patients may be more likely to complete a paper one.⁶⁷ Patients prefer electronic diaries over long paper forms,⁶⁸ but a practical issue to consider is easy electronic access.

Patients can start keeping a headache diary before the initial consultation to assist with diagnosis, or early in their management. A first-appointment diary mailed with instructions is a feasible option.⁶⁹ These types of diaries ask detailed questions to help diagnose all major primary headache types including menstrual migraine and to identify concomitant medication-overuse headache. Physicians and patients generally report improved communication with use of a diary.⁷⁰

Some providers distinguish between a headache diary and a calendar. In standard practice, a headache diary is the general term referring to both, but the literature differentiates between the two. Both should at least include headache frequency, with possible inclusion of other factors such as headache duration, headache intensity, analgesic use, headache impact on function, and absenteeism. Potential triggers including menses can also be tracked. The calendar version can fit on a single page and can be used for simple tracking of headache frequency and analgesia use.

One of the simplest calendars to use is the “stoplight” calendar. Red days are when a patient is completely debilitated in bed. On a yellow day, function at work, school, or daily activities is significantly reduced by migraine, but the patient is not bedbound. A green day is when headache is present but function is not affected. No color is placed if the patient is 100% headache-free.

Acute treatment use can be written in or, to improve compliance, a checkmark can be placed on days of treatment. Patients who are tracking menses circle the days of menstruation. The calendar-diary should be brought to every appointment to track treatment response and medication use.

THE SECOND S IS FOR STRESS

Behavioral management such as cognitive behavioral therapy in migraine has been shown to decrease catastrophizing, migraine disability, and headache severity and frequency.⁷⁴ Both depression and anxiety can improve along with migraine.⁷⁵ Cognitive behavioral therapy can be provided in individualized sessions or group sessions, either in person or online.^74,76,77 The effects become more prominent about 5 weeks into treatment.⁷⁸

Biofeedback, which uses behavioral techniques paired with physiologic autonomic measures, has been extensively studied, and shows benefit in migraine, including in meta-analysis.⁷⁹ The types of biofeedback measurements used include electromyography, electroencephalography, temperature, sweat sensors, heart rate, blood volume pulse feedback, and respiration bands. While biofeedback is generally done under the guidance of a therapist, it can still be useful with minimal therapist contact and supplemental audio.⁸⁰

Mindfulness, or the awareness of thoughts, feelings, and sensations in the present moment without judgment, is a behavioral technique that can be done alone or paired with another technique. It is often taught through a mindfulness-based stress-reduction  program, which relies on a standardized approach. A meta-analysis showed that mindfulness improves pain intensity, headache frequency, disability, self-efficacy, and quality of life.⁸¹ It may work by encouraging pain acceptance.⁸²

Relaxation techniques are also employed in migraine management, either alone or in conjunction with techniques mentioned  above, such as mindfulness. They include progressive muscle relaxation and deep breathing. Relaxation has been shown to be effective when done by professional trainers as well as lay trainers in both individual and group settings.^83,84

In patients with intractable headache, more-intensive inpatient and outpatient programs have been tried. Inpatient admissions with multidisciplinary programs that include a focus on behavioral techniques often paired with lifestyle education and sometimes pharmacologic management can be beneficial.^85,86 These programs have also been successfully conducted as multiple outpatient sessions.^86–88

Stress management is an important aspect of migraine management. These treatments often involve homework and require active participation.

LIFESTYLE FOR ALL

All patients with migraine should initiate lifestyle modifications (see Advice to patients with migraine: SEEDS for success). Modifications with the highest level of evidence, specifically behavioral techniques, have had the most reproducible results. A headache diary is an essential tool to identify patterns and needs for optimization of acute or preventive treatment regimens. The strongest evidence is for the behavioral management techniques for stress reduction.

Migraine is the second leading cause of years of life lived with a disability globally.¹ It affects people of all ages, but particularly during the years associated with the highest productivity in terms of work and family life.

Migraine is a genetic neurologic disease that can be influenced or triggered by environmental factors. However, triggers do not cause migraine. For example, stress does not cause migraine, but it can exacerbate it.

Primary care physicians can help patients reduce the likelihood of a migraine attack, the severity of symptoms, or both by offering lifestyle counseling centered around the mnemonic SEEDS: sleep, exercise, eat, diary, and stress. In this article, each factor is discussed individually for its current support in the literature along with best-practice recommendations.

S IS FOR SLEEP

Before optimizing sleep hygiene, screen for sleep apnea, especially in those who have chronic daily headache upon awakening. An excellent tool is the STOP-Bang screening questionnaire⁵ (www.stopbang.ca/osa/screening.php). Patients respond “yes” or “no” to the following questions:

• Snoring: Do you snore loudly (louder than talking or loud enough to be heard through closed doors)?
• Tired: Do you often feel tired, fatigued, or sleepy during the daytime?
• Observed: Has anyone observed you stop breathing during your sleep?
• Pressure: Do you have or are you being treated for high blood pressure?
• Body mass index greater than 35 kg/m²?
• Age over 50?
• Neck circumference larger than 40 cm (females) or  42 cm (males)?
• Gender—male?

Each “yes” answer is scored as 1 point. A score less than 3 indicates low risk of obstructive sleep apnea; 3 to 4 indicates moderate risk; and 5 or more indicates high risk. Optimization of sleep apnea with continuous positive airway pressure therapy can improve sleep apnea headache.⁶ The improved sleep from reduced arousals may also mitigate migraine symptoms.

Behavioral modification for sleep hygiene can convert chronic migraine to episodic migraine.⁷ One such program is stimulus control therapy, which focuses on using cues to initiate sleep (Table 1). Patients are encouraged to keep the bedroom quiet, dark, and cool, and to go to sleep at the same time every night. Importantly, the bed should be associated only with sleep. If patients are unable to fall asleep within 20 to 30 minutes, they should leave the room so they do not associate the bed with frustration and anxiety. Use of phones, tablets, and television in the bedroom is discouraged as these devices may make it more difficult to fall asleep.⁸

The next option is sleep restriction, which is useful for comorbid insomnia. Patients keep a sleep diary to better understand their sleep-wake cycle. The goal is 90% sleep efficiency, meaning that 90% of the time in bed (TIB) is spent asleep. For example, if the patient is in bed 8 hours but asleep only 4 hours, sleep efficiency is 50%. The goal is to reduce TIB to match the time asleep and to agree on a prescribed daily wake-up time. When the patient is consistently sleeping 90% of the TIB, add 30-minute increments until he or she is appropriately sleeping 7 to 8 hours at night.⁹ Naps are not recommended.

Let patients know that their migraine may worsen until a new routine sleep pattern emerges. This method is not recommended for patients with untreated sleep apnea.

E IS FOR EXERCISE

Exercise is broadly recommended for a healthy lifestyle; some evidence suggests that it can also be useful in the management of migraine.¹⁰ Low levels of physical activity and a sedentary lifestyle are associated with migraine.¹¹ It is unclear if patients with migraine are less likely to exercise because they want to avoid triggering a migraine or if a sedentary lifestyle increases their risk.

Exercise has been studied for its prophylactic benefits in migraine, and one hypothesis relates to beta-endorphins. Levels of beta-endorphins are reduced in the cerebrospinal fluid of patients with migraine.¹² Exercise programs may increase levels while reducing headache frequency and duration.¹³ One study showed that pain thresholds do not change with exercise programs, suggesting that it is avoidance behavior that is positively altered rather than the underlying pain pathways.¹⁴

A systematic review and meta-analysis based on 5 randomized controlled trials and 1 nonrandomized controlled clinical trial showed that exercise reduced monthly migraine days by only 0.6 (± 0.3) days, but the data also suggested that as the exercise intensity increased, so did the positive effects.¹⁰

Some data suggest that exercise may also reduce migraine duration and severity as well as the need for abortive medication.¹⁰ Two studies in this systematic review^15,16 showed that exercise benefits were equivalent to those of migraine preventives such as amitriptyline and topiramate; the combination of amitriptyline and exercise was more beneficial than exercise alone. Multiple types of exercise were beneficial, including walking, jogging, cross-training, and cycling when done for least 6 weeks and for 30 to 50 minutes 3 to 5 times a week.

These findings are in line with the current recommendations for general health from the American College of Sports Medicine, ie, moderate to vigorous cardio­respiratory exercise for 30 to 60 minutes 3 to 5 times a week (or 150 minutes per week). The daily exercise can be continuous or done in intervals of less than 20 minutes. For those with a sedentary lifestyle, as is seen in a significant proportion of the migraine population, light to moderate exercise for less than 20 minutes is still beneficial.¹⁷

Based on this evidence, the best current recommendation for patients with migraine is to engage in graded moderate cardiorespiratory exercise, although any exercise is better than none. If a patient is sedentary or has poor exercise tolerance, or both, exercising once a week for shorter time periods may be a manageable place to start.

Some patients may identify exercise as a trigger or exacerbating factor in migraine. These patients may need appropriate prophylactic and abortive therapies before starting an exercise regimen.

THE SECOND E IS FOR EAT (FOOD AND DRINK)

Many patients believe that some foods trigger migraine attacks, but further study is needed. The most consistent food triggers appear to be red wine and caffeine (withdrawal).^18,19 Interestingly, patients with migraine report low levels of alcohol consumption,²⁰ but it is unclear if that is because alcohol has a protective effect or if patients avoid it.

Some patients may crave certain foods in the prodromal phase of an attack, eat the food, experience the attack, and falsely conclude that the food caused the attack.²¹ Premonitory symptoms include fatigue, cognitive changes, homeostatic changes, sensory hyperresponsiveness, and food cravings.²¹ It is difficult to distinguish between premonitory phase food cravings and true triggers because premonitory symptoms can precede headache by 48 to 72 hours, and the timing for a trigger to be considered causal is not known.²²

Chocolate is often thought to be a migraine trigger, but the evidence argues against this and even suggests that sweet cravings are a part of the premonitory phase.²³ Monosodium glutamate is often identified as a trigger as well, but the literature is inconsistent and does not support a causal relationship.²⁴ Identifying true food triggers in migraine is difficult, and patients with migraine may have poor quality diets, with some foods acting as true triggers for certain patients.²⁵ These possibilities have led to the development of many “migraine diets,” including elimination diets.

Elimination diets

Elimination diets involve avoiding specific food items over a period of time and then adding them back in one at a time to gauge whether they cause a reaction in the body. A number of these diets have been studied for their effects on headache and migraine:

Gluten-free diets restrict foods that contain wheat, rye, and barley. A systematic review of gluten-free diets in patients with celiac disease found that headache or migraine frequency decreased by 51.6% to 100% based on multiple cohort studies (N = 42,388).²⁶ There are no studies on the use of a gluten-free diet for migraine in patients without celiac disease.

Immunoglobulin G-elimination diets restrict foods that serve as antigens for IgG. However, data supporting these diets are inconsistent. Two small randomized controlled trials found that the diets improved migraine symptoms, but a larger study found no improvement in the number of migraine days at 12 weeks, although there was an initially significant effect at 4 weeks.^27–29

Antihistamine diets restrict foods that have high levels of histamines, including fermented dairy, vegetables, soy products,  wine, beer, alcohol, and those that cause histamine release regardless of IgE testing results. A prospective single-arm study of antihistamine diets in patients with chronic headache reported symptom improvement, which could be applied to certain comorbidities such as  mast cell activation syndrome.³⁰ Another prospective nonrandomized controlled study eliminated foods based on positive IgE skin-prick testing for allergy in patients with recurrent migraine and found that it reduced headache frequency.³¹

Tyramine-free diets are often recommended due to the presumption that tyramine-containing foods (eg, aged cheese, cured or smoked meats and fish, and beer) are triggers. However, multiple studies have reviewed this theory with inconsistent results,³² and the only study of a tyramine-free diet was negative.³³ In addition, commonly purported high-tyramine foods have lower tyramine levels than previously thought.³⁴

Low-fat diets in migraine are supported by 2 small randomized controlled trials and a prospective study showing a decrease in symptom severity; the results for frequency are inconsistent.^35–37

Low-glycemic index diets are supported in migraine by 1 randomized controlled trial that showed improvement in migraine frequency in a diet group and in a control group of patients who took a standard migraine-preventive medication to manage their symptoms.³⁸

Other migraine diets

Diets high in certain foods or ingredient ratios, as opposed to elimination diets, have also been studied in patients with migraine. One promising diet containing high levels of omega-3 fatty acids and low levels of omega-6 fatty acids was shown in a systematic review to reduce the duration of migraine but not the frequency or severity.³⁹ A more recent randomized controlled trial of this diet in chronic migraine also showed that it decreased migraine frequency.⁴⁰

The ketogenic diet (high fat, low carbohydrate) had promising results in a randomized controlled trial in overweight women with migraine and in a prospective study.^41,42 However, a prospective study of the Atkins diet in teenagers with chronic daily headaches showed no benefit.⁴³ The ketogenic diet is difficult to follow and may work in part due to weight loss alone, although ketogenesis itself may also play a role.^41,44

Sodium levels have been shown to be higher in the cerebrospinal fluid of patients with migraine than in controls, particularly during an attack.⁴⁵ For a prehypertensive population or an elderly population, a low-sodium diet may be beneficial based on 2 prospective trials.^46,47 However, a younger female population without hypertension and low-to-normal body mass index had a reduced probability of migraine while consuming a high-sodium diet.⁴⁸

Counseling about sodium intake should be tailored to specific patient populations. For example, a diet low in sodium may be appropriate for patients with vascular risk factors such as hypertension, whereas a high-sodium diet may be appropriate in patients with comorbidities like postural tachycardia syndrome or in those with a propensity for low blood pressure or low body mass index.

Encourage routine meals and hydration

The standard advice for patients with migraine is to consume regular meals. Headaches have been associated with fasting, and those with migraine are predisposed to attacks in the setting of fasting.^49,50 Migraine is more common when meals are skipped, particularly breakfast.⁵¹

It is unclear how fasting lowers the migraine threshold. Nutritional studies show that skipping meals, particularly breakfast, increases low-grade inflammation and impairs  glucose metabolism by affecting insulin and fat oxidation metabolism.⁵² However, hypoglycemia itself is not a consistent cause of headache or migraine attacks.⁵³ As described above, a randomized controlled trial of a low-glycemic index diet actually decreased migraine frequency and severity.³⁸ Skipping meals also reduces energy and is associated with reduced physical activity, perhaps leading to multiple compounding triggers that further lower the migraine threshold.^54,55

When counseling patients about the need to eat breakfast, consider what they normally consume (eg, is breakfast just a cup of coffee?). Replacing simple carbohydrates with protein, fats, and fiber may be beneficial for general health, but the effects on migraine are not known, nor is the optimal composition of breakfast foods.⁵⁵

The optimal timing of breakfast relative to awakening is also unclear, but in general, it should be eaten within 30 to 60 minutes of rising. Also consider patients’ work hours—delayed-phase or shift workers have altered sleep cycles.

Recommendations vary in regard to hydration. Headache is associated with fluid restriction and dehydration,^56,57 but only a few studies suggest that rehydration and increased hydration status can improve migraine.⁵⁸ In fact, a single post hoc analysis of a metoclopramide study showed that intravenous fluid alone for patients with migraine in the emergency room did not improve pain outcomes.⁵⁹

The amount of water patients should drink daily in the setting of migraine is also unknown, but a study showed benefit with 4 L, which equates to a daily intake of 16 eight-ounce glasses.⁶⁰ One review on general health that could be extrapolated given the low risk of the intervention indicated that 1.8 L daily (7 to 8 eight-ounce glasses) promoted a euhydration status in most people, although many factors contribute to hydration status.⁶¹

Caffeine intake is also a major consideration. Caffeine is a nonspecific adenosine receptor antagonist that modulates adenosine receptors like the pronociceptive 2A receptor, leading to changes integral to the neuropathophysiology of migraine.⁶² Caffeine has analgesic properties at doses greater than 65 to 200 mg and augments the effects of analgesics such as acetaminophen and aspirin. Chronic caffeine use can lead to withdrawal symptoms when intake is stopped abruptly; this is thought to be due to upregulation of adenosine receptors, but the effect varies based on genetic predisposition.¹⁹

The risk of chronic daily headache may relate to high use of caffeine preceding the onset of chronification, and caffeine abstinence may improve response to acute migraine treatment.^19,63 There is a dose-dependent risk of headache.^64,65 Current recommendations suggest limiting caffeine consumption to less than 200 mg per day or stopping caffeine consumption altogether based on the quantity required for caffeine-withdrawal headache.⁶⁶ Varying  the caffeine dose from day to day may also trigger headache due to the high sensitivity to caffeine withdrawal.

While many diets have shown potential benefit in patients with migraine, more studies are needed before any one “migraine diet” can be recommended. Caution should be taken, as there is risk of adverse effects from nutrient deficiencies or excess levels, especially if the patient is not under the care of a healthcare professional who is familiar with the diet.

Whether it is beneficial to avoid specific food triggers at this time is unclear and still controversial even within the migraine community because some of these foods may be misattributed as triggers instead of premonitory cravings driven by the hypothalamus. It is important to counsel patients with migraine to eat a healthy diet with consistent meals, to maintain adequate hydration, and to keep their caffeine intake low or at least consistent, although these teachings are predominantly based on limited studies with extrapolation from nutrition research.

D IS FOR DIARY

A headache diary is a recommended part of headache management and may enhance the accuracy of diagnosis and assist in treatment modifications. Paper and electronic diaries have been used. Electronic diaries may be more accurate for real-time use, but patients may be more likely to complete a paper one.⁶⁷ Patients prefer electronic diaries over long paper forms,⁶⁸ but a practical issue to consider is easy electronic access.

Patients can start keeping a headache diary before the initial consultation to assist with diagnosis, or early in their management. A first-appointment diary mailed with instructions is a feasible option.⁶⁹ These types of diaries ask detailed questions to help diagnose all major primary headache types including menstrual migraine and to identify concomitant medication-overuse headache. Physicians and patients generally report improved communication with use of a diary.⁷⁰

Some providers distinguish between a headache diary and a calendar. In standard practice, a headache diary is the general term referring to both, but the literature differentiates between the two. Both should at least include headache frequency, with possible inclusion of other factors such as headache duration, headache intensity, analgesic use, headache impact on function, and absenteeism. Potential triggers including menses can also be tracked. The calendar version can fit on a single page and can be used for simple tracking of headache frequency and analgesia use.

One of the simplest calendars to use is the “stoplight” calendar. Red days are when a patient is completely debilitated in bed. On a yellow day, function at work, school, or daily activities is significantly reduced by migraine, but the patient is not bedbound. A green day is when headache is present but function is not affected. No color is placed if the patient is 100% headache-free.

Acute treatment use can be written in or, to improve compliance, a checkmark can be placed on days of treatment. Patients who are tracking menses circle the days of menstruation. The calendar-diary should be brought to every appointment to track treatment response and medication use.

THE SECOND S IS FOR STRESS

Behavioral management such as cognitive behavioral therapy in migraine has been shown to decrease catastrophizing, migraine disability, and headache severity and frequency.⁷⁴ Both depression and anxiety can improve along with migraine.⁷⁵ Cognitive behavioral therapy can be provided in individualized sessions or group sessions, either in person or online.^74,76,77 The effects become more prominent about 5 weeks into treatment.⁷⁸

Biofeedback, which uses behavioral techniques paired with physiologic autonomic measures, has been extensively studied, and shows benefit in migraine, including in meta-analysis.⁷⁹ The types of biofeedback measurements used include electromyography, electroencephalography, temperature, sweat sensors, heart rate, blood volume pulse feedback, and respiration bands. While biofeedback is generally done under the guidance of a therapist, it can still be useful with minimal therapist contact and supplemental audio.⁸⁰

Mindfulness, or the awareness of thoughts, feelings, and sensations in the present moment without judgment, is a behavioral technique that can be done alone or paired with another technique. It is often taught through a mindfulness-based stress-reduction  program, which relies on a standardized approach. A meta-analysis showed that mindfulness improves pain intensity, headache frequency, disability, self-efficacy, and quality of life.⁸¹ It may work by encouraging pain acceptance.⁸²

Relaxation techniques are also employed in migraine management, either alone or in conjunction with techniques mentioned  above, such as mindfulness. They include progressive muscle relaxation and deep breathing. Relaxation has been shown to be effective when done by professional trainers as well as lay trainers in both individual and group settings.^83,84

In patients with intractable headache, more-intensive inpatient and outpatient programs have been tried. Inpatient admissions with multidisciplinary programs that include a focus on behavioral techniques often paired with lifestyle education and sometimes pharmacologic management can be beneficial.^85,86 These programs have also been successfully conducted as multiple outpatient sessions.^86–88

Stress management is an important aspect of migraine management. These treatments often involve homework and require active participation.

LIFESTYLE FOR ALL

All patients with migraine should initiate lifestyle modifications (see Advice to patients with migraine: SEEDS for success). Modifications with the highest level of evidence, specifically behavioral techniques, have had the most reproducible results. A headache diary is an essential tool to identify patterns and needs for optimization of acute or preventive treatment regimens. The strongest evidence is for the behavioral management techniques for stress reduction.

Migraine is the second leading cause of years of life lived with a disability globally.¹ It affects people of all ages, but particularly during the years associated with the highest productivity in terms of work and family life.

Migraine is a genetic neurologic disease that can be influenced or triggered by environmental factors. However, triggers do not cause migraine. For example, stress does not cause migraine, but it can exacerbate it.

Primary care physicians can help patients reduce the likelihood of a migraine attack, the severity of symptoms, or both by offering lifestyle counseling centered around the mnemonic SEEDS: sleep, exercise, eat, diary, and stress. In this article, each factor is discussed individually for its current support in the literature along with best-practice recommendations.

S IS FOR SLEEP

Before optimizing sleep hygiene, screen for sleep apnea, especially in those who have chronic daily headache upon awakening. An excellent tool is the STOP-Bang screening questionnaire⁵ (www.stopbang.ca/osa/screening.php). Patients respond “yes” or “no” to the following questions:

• Snoring: Do you snore loudly (louder than talking or loud enough to be heard through closed doors)?
• Tired: Do you often feel tired, fatigued, or sleepy during the daytime?
• Observed: Has anyone observed you stop breathing during your sleep?
• Pressure: Do you have or are you being treated for high blood pressure?
• Body mass index greater than 35 kg/m²?
• Age over 50?
• Neck circumference larger than 40 cm (females) or  42 cm (males)?
• Gender—male?

Each “yes” answer is scored as 1 point. A score less than 3 indicates low risk of obstructive sleep apnea; 3 to 4 indicates moderate risk; and 5 or more indicates high risk. Optimization of sleep apnea with continuous positive airway pressure therapy can improve sleep apnea headache.⁶ The improved sleep from reduced arousals may also mitigate migraine symptoms.

Behavioral modification for sleep hygiene can convert chronic migraine to episodic migraine.⁷ One such program is stimulus control therapy, which focuses on using cues to initiate sleep (Table 1). Patients are encouraged to keep the bedroom quiet, dark, and cool, and to go to sleep at the same time every night. Importantly, the bed should be associated only with sleep. If patients are unable to fall asleep within 20 to 30 minutes, they should leave the room so they do not associate the bed with frustration and anxiety. Use of phones, tablets, and television in the bedroom is discouraged as these devices may make it more difficult to fall asleep.⁸

The next option is sleep restriction, which is useful for comorbid insomnia. Patients keep a sleep diary to better understand their sleep-wake cycle. The goal is 90% sleep efficiency, meaning that 90% of the time in bed (TIB) is spent asleep. For example, if the patient is in bed 8 hours but asleep only 4 hours, sleep efficiency is 50%. The goal is to reduce TIB to match the time asleep and to agree on a prescribed daily wake-up time. When the patient is consistently sleeping 90% of the TIB, add 30-minute increments until he or she is appropriately sleeping 7 to 8 hours at night.⁹ Naps are not recommended.

Let patients know that their migraine may worsen until a new routine sleep pattern emerges. This method is not recommended for patients with untreated sleep apnea.

E IS FOR EXERCISE

Exercise is broadly recommended for a healthy lifestyle; some evidence suggests that it can also be useful in the management of migraine.¹⁰ Low levels of physical activity and a sedentary lifestyle are associated with migraine.¹¹ It is unclear if patients with migraine are less likely to exercise because they want to avoid triggering a migraine or if a sedentary lifestyle increases their risk.

Exercise has been studied for its prophylactic benefits in migraine, and one hypothesis relates to beta-endorphins. Levels of beta-endorphins are reduced in the cerebrospinal fluid of patients with migraine.¹² Exercise programs may increase levels while reducing headache frequency and duration.¹³ One study showed that pain thresholds do not change with exercise programs, suggesting that it is avoidance behavior that is positively altered rather than the underlying pain pathways.¹⁴

A systematic review and meta-analysis based on 5 randomized controlled trials and 1 nonrandomized controlled clinical trial showed that exercise reduced monthly migraine days by only 0.6 (± 0.3) days, but the data also suggested that as the exercise intensity increased, so did the positive effects.¹⁰

Some data suggest that exercise may also reduce migraine duration and severity as well as the need for abortive medication.¹⁰ Two studies in this systematic review^15,16 showed that exercise benefits were equivalent to those of migraine preventives such as amitriptyline and topiramate; the combination of amitriptyline and exercise was more beneficial than exercise alone. Multiple types of exercise were beneficial, including walking, jogging, cross-training, and cycling when done for least 6 weeks and for 30 to 50 minutes 3 to 5 times a week.

These findings are in line with the current recommendations for general health from the American College of Sports Medicine, ie, moderate to vigorous cardio­respiratory exercise for 30 to 60 minutes 3 to 5 times a week (or 150 minutes per week). The daily exercise can be continuous or done in intervals of less than 20 minutes. For those with a sedentary lifestyle, as is seen in a significant proportion of the migraine population, light to moderate exercise for less than 20 minutes is still beneficial.¹⁷

Based on this evidence, the best current recommendation for patients with migraine is to engage in graded moderate cardiorespiratory exercise, although any exercise is better than none. If a patient is sedentary or has poor exercise tolerance, or both, exercising once a week for shorter time periods may be a manageable place to start.

Some patients may identify exercise as a trigger or exacerbating factor in migraine. These patients may need appropriate prophylactic and abortive therapies before starting an exercise regimen.

THE SECOND E IS FOR EAT (FOOD AND DRINK)

Many patients believe that some foods trigger migraine attacks, but further study is needed. The most consistent food triggers appear to be red wine and caffeine (withdrawal).^18,19 Interestingly, patients with migraine report low levels of alcohol consumption,²⁰ but it is unclear if that is because alcohol has a protective effect or if patients avoid it.

Some patients may crave certain foods in the prodromal phase of an attack, eat the food, experience the attack, and falsely conclude that the food caused the attack.²¹ Premonitory symptoms include fatigue, cognitive changes, homeostatic changes, sensory hyperresponsiveness, and food cravings.²¹ It is difficult to distinguish between premonitory phase food cravings and true triggers because premonitory symptoms can precede headache by 48 to 72 hours, and the timing for a trigger to be considered causal is not known.²²

Chocolate is often thought to be a migraine trigger, but the evidence argues against this and even suggests that sweet cravings are a part of the premonitory phase.²³ Monosodium glutamate is often identified as a trigger as well, but the literature is inconsistent and does not support a causal relationship.²⁴ Identifying true food triggers in migraine is difficult, and patients with migraine may have poor quality diets, with some foods acting as true triggers for certain patients.²⁵ These possibilities have led to the development of many “migraine diets,” including elimination diets.

Elimination diets

Elimination diets involve avoiding specific food items over a period of time and then adding them back in one at a time to gauge whether they cause a reaction in the body. A number of these diets have been studied for their effects on headache and migraine:

Gluten-free diets restrict foods that contain wheat, rye, and barley. A systematic review of gluten-free diets in patients with celiac disease found that headache or migraine frequency decreased by 51.6% to 100% based on multiple cohort studies (N = 42,388).²⁶ There are no studies on the use of a gluten-free diet for migraine in patients without celiac disease.

Immunoglobulin G-elimination diets restrict foods that serve as antigens for IgG. However, data supporting these diets are inconsistent. Two small randomized controlled trials found that the diets improved migraine symptoms, but a larger study found no improvement in the number of migraine days at 12 weeks, although there was an initially significant effect at 4 weeks.^27–29

Antihistamine diets restrict foods that have high levels of histamines, including fermented dairy, vegetables, soy products,  wine, beer, alcohol, and those that cause histamine release regardless of IgE testing results. A prospective single-arm study of antihistamine diets in patients with chronic headache reported symptom improvement, which could be applied to certain comorbidities such as  mast cell activation syndrome.³⁰ Another prospective nonrandomized controlled study eliminated foods based on positive IgE skin-prick testing for allergy in patients with recurrent migraine and found that it reduced headache frequency.³¹

Tyramine-free diets are often recommended due to the presumption that tyramine-containing foods (eg, aged cheese, cured or smoked meats and fish, and beer) are triggers. However, multiple studies have reviewed this theory with inconsistent results,³² and the only study of a tyramine-free diet was negative.³³ In addition, commonly purported high-tyramine foods have lower tyramine levels than previously thought.³⁴

Low-fat diets in migraine are supported by 2 small randomized controlled trials and a prospective study showing a decrease in symptom severity; the results for frequency are inconsistent.^35–37

Low-glycemic index diets are supported in migraine by 1 randomized controlled trial that showed improvement in migraine frequency in a diet group and in a control group of patients who took a standard migraine-preventive medication to manage their symptoms.³⁸

Other migraine diets

Diets high in certain foods or ingredient ratios, as opposed to elimination diets, have also been studied in patients with migraine. One promising diet containing high levels of omega-3 fatty acids and low levels of omega-6 fatty acids was shown in a systematic review to reduce the duration of migraine but not the frequency or severity.³⁹ A more recent randomized controlled trial of this diet in chronic migraine also showed that it decreased migraine frequency.⁴⁰

The ketogenic diet (high fat, low carbohydrate) had promising results in a randomized controlled trial in overweight women with migraine and in a prospective study.^41,42 However, a prospective study of the Atkins diet in teenagers with chronic daily headaches showed no benefit.⁴³ The ketogenic diet is difficult to follow and may work in part due to weight loss alone, although ketogenesis itself may also play a role.^41,44

Sodium levels have been shown to be higher in the cerebrospinal fluid of patients with migraine than in controls, particularly during an attack.⁴⁵ For a prehypertensive population or an elderly population, a low-sodium diet may be beneficial based on 2 prospective trials.^46,47 However, a younger female population without hypertension and low-to-normal body mass index had a reduced probability of migraine while consuming a high-sodium diet.⁴⁸

Counseling about sodium intake should be tailored to specific patient populations. For example, a diet low in sodium may be appropriate for patients with vascular risk factors such as hypertension, whereas a high-sodium diet may be appropriate in patients with comorbidities like postural tachycardia syndrome or in those with a propensity for low blood pressure or low body mass index.

Encourage routine meals and hydration

The standard advice for patients with migraine is to consume regular meals. Headaches have been associated with fasting, and those with migraine are predisposed to attacks in the setting of fasting.^49,50 Migraine is more common when meals are skipped, particularly breakfast.⁵¹

It is unclear how fasting lowers the migraine threshold. Nutritional studies show that skipping meals, particularly breakfast, increases low-grade inflammation and impairs  glucose metabolism by affecting insulin and fat oxidation metabolism.⁵² However, hypoglycemia itself is not a consistent cause of headache or migraine attacks.⁵³ As described above, a randomized controlled trial of a low-glycemic index diet actually decreased migraine frequency and severity.³⁸ Skipping meals also reduces energy and is associated with reduced physical activity, perhaps leading to multiple compounding triggers that further lower the migraine threshold.^54,55

When counseling patients about the need to eat breakfast, consider what they normally consume (eg, is breakfast just a cup of coffee?). Replacing simple carbohydrates with protein, fats, and fiber may be beneficial for general health, but the effects on migraine are not known, nor is the optimal composition of breakfast foods.⁵⁵

The optimal timing of breakfast relative to awakening is also unclear, but in general, it should be eaten within 30 to 60 minutes of rising. Also consider patients’ work hours—delayed-phase or shift workers have altered sleep cycles.

Recommendations vary in regard to hydration. Headache is associated with fluid restriction and dehydration,^56,57 but only a few studies suggest that rehydration and increased hydration status can improve migraine.⁵⁸ In fact, a single post hoc analysis of a metoclopramide study showed that intravenous fluid alone for patients with migraine in the emergency room did not improve pain outcomes.⁵⁹

The amount of water patients should drink daily in the setting of migraine is also unknown, but a study showed benefit with 4 L, which equates to a daily intake of 16 eight-ounce glasses.⁶⁰ One review on general health that could be extrapolated given the low risk of the intervention indicated that 1.8 L daily (7 to 8 eight-ounce glasses) promoted a euhydration status in most people, although many factors contribute to hydration status.⁶¹