Clinical Review

Dr. Kaunitz receives grant or research support from Bayer, Agile, Noven, Teva, and Medical Diagnostic Laboratories, is a consultant to Bayer, Merck, and Teva, and owns stock in Becton Dickinson.

Among the developments of the past year in the care of menopausal women are:

• updated guidelines from the Institute of Medicine regarding vitamin D requirements—suggesting that fewer women are deficient in this nutrient than experts had believed
• new data from Europe on hormone therapy (HT) that highlight the safety of transdermal estrogen in comparison with oral administration
• a recent analysis from the Women’s Health Initiative (WHI), confirming a small elevated risk of breast cancer mortality with use of combination estrogen-progestin HT
• confirmation that age at initiation of HT determines its effect on cardiovascular health
• clarification of the association between HT and dementia
• new data demonstrating modest improvement in hot flushes when the serotonin reuptake inhibitor (SRI) escitalopram is used
• a brand new report from the WHI estrogen-alone arm that shows a protective effect against breast cancer.

The new data on HT suggest that we still have much to learn about its benefits and risks. We also are reaching an understanding that, for many young, symptomatic, menopausal patients, HT can represent a safe choice, with much depending on the timing and duration of therapy.

For more on how your colleagues are managing menopausal patients with and without hormone therapy, see “Is hormone therapy still a valid option? 12 ObGyns address this question,” on the facing page.

Menopausal women need less vitamin D than we thought

Institute of Medicine. Dietary reference intakes for calcium and vitamin D. Washington, DC: IOM; December 2010. http://www.iom.edu/~/media/Files/Report%20Files/2010/Dietary-Reference-Intakes-for-Calcium-and-Vitamin-D/Vitamin%20D%20and%20Calcium%202010%20Report%20Brief.pdf. Accessed March 24, 2011.

In the 2010 Update on Menopause, I summarized recent findings on vitamin D requirements, including recommendations that menopausal women should take at least 800 IU of vitamin D daily. I also described the prevailing expert opinion that many North American women are deficient in this nutrient.

What a difference a year can make! In late November, the Institute of Medicine (IOM) released a comprehensive report on vitamin D. Here are some of its conclusions:

• Vitamin D plays an important role in skeletal health but its role in other areas, including cardiovascular disease and cancer, is uncertain
• An intake of 600 IU of vitamin D daily is appropriate for girls and for women as old as 70 years; an in-take of 800 IU daily is appropriate for women older than 70 years
• A serum level of 25-hydroxy vitamin D of 20 ng/mL is consistent with adequate vitamin D status; this is lower than the threshold many have recommended
• With few exceptions, all people who live in North America—including those who have minimal or no exposure to sunlight—are receiving adequate calcium and vitamin D
• Ingestion of more than 4,000 IU of vitamin D daily can cause renal damage and injure other tissues.

The IOM report will likely prompt multivitamin manufacturers to increase the amount of vitamin D contained in their supplements to 600 IU daily. In addition, the report will probably discourage the common practice of checking serum 25-hydroxy vitamin D levels and prescribing a high dosage of vitamin D supplementation when the level is below 30 ng/mL.

Is transdermal estrogen safer than oral administration?

Canonico M, Fournier A, Carcaillon L, et al. Postmenopausal hormone therapy and risk of idiopathic venous thromboembolism: results from the E3N cohort study. Arterioscler Thromb Vasc Biol. 2010;30(2):340–345.

Renoux C, Dell’aniello S, Garbe E, Suissa S. Transdermal and oral hormone replacement therapy and the risk of stroke: a nested case-control study. BMJ. 2010;340:c2519. doi: 10.1136/bmj.c2519.

In the WHI, the combination of oral conjugated equine estrogen and medroxyprogesterone acetate more than doubled the risk of deep venous thrombosis and pulmonary embolism and modestly increased the risk of stroke, compared with nonuse.¹

A year after publication of the initial findings of the WHI estrogen-progestin arm, the Estrogen and THromboEmbolism Risk Study Group (ESTHER) case-control study from France provided evidence that transdermal estrogen does not increase the risk of venous thrombosis.² In France, many menopausal women use HT, and the transdermal route of administration is common.

In 2010, the E3N cohort study from France also assessed the risk of thrombosis associated with oral and transdermal HT. Investigators followed more than 80,000 postmenopausal women and found that, unlike oral HT, the transdermal route did not increase the risk of venous thrombosis.

More recent evidence also suggests a safety advantage for transdermal HT. The newest data come from the United Kingdom General Practice Research Database, which includes information on more than 870,000 women who were 50 to 70 years old from 1987 to 2006. Investigators identified more than 15,000 women who were given a diagnosis of stroke during this period and compared the use of HT in these women with that of almost 60,000 women in a control group. The risk of stroke associated with current use of transdermal HT was similar to the risk associated with nonuse of HT. Women who used a patch containing 0.05 mg of estradiol or less had a risk of stroke 19% lower than women who did not use HT.

In contrast, the risk of stroke in users of patches that contained a higher dosage of estradiol was almost twice the risk in nonusers of HT. Current users of oral HT had a risk of stroke 28% higher than that of nonusers of HT.

Estrogen-progestin HT raises the risk of death from breast cancer

Chlebowski RT, Anderson GL, Gass M, et al. Estrogen plus progestin and breast cancer incidence and mortality in postmenopausal women. JAMA. 2010;304(15):1684–1692.

Toh S, Hernandez-Diaz S, Logan R, Rossouw JE, Hernan MA. Coronary heart disease in postmenopausal recipients of estrogen plus progestin: does the increased risk ever disappear? Ann Intern Med. 2010;152(4):211–217.

In the estrogen-progestin arm of the WHI, initially published in 2002, the risk of invasive breast cancer was modestly elevated (hazard ratio [HR], 1.26) among women who had used HT longer than 5 years.³

In 2010, investigators reported on breast cancer mortality in WHI participants at a mean follow-up of 11 years. They found that combination HT users had breast cancer histology similar to that of nonusers. However, the tumors were more likely to be node-positive in combination HT users (23.7% vs 16.2%). In addition, breast cancer mortality was slightly higher among users of HT (2.6 vs 1.3 deaths in every 10,000 woman-years) (HR, 1.96; 95% confidence interval, 1.00–4.04).

Earlier observational studies had suggested that the death rate from breast cancer is lower in users of combination HT than in nonusers. Consistent with the UK Million Women Study, however, a 2010 report from the WHI found a higher mortality rate among women who have used HT.⁴

These new WHI findings reinforce the importance of assessing whether micronized progesterone combined with estrogen might lower the risk of death from breast cancer—a possibility suggested by findings of the French E3N cohort study.⁵

In addition, given the possibility that HT may be cardioprotective when it is initiated within 10 years after the onset of menopause, a WHI report that addresses long-term all-cause mortality would allow us to better counsel our menopausal patients who are trying to decide whether to start or continue HT. See, for example, the data from the California Teachers Study (below) and the estrogen-alone arm of the WHI (page 46).

Age at initiation of HT determines its effect on CHD

Stram DO, Liu Y, Henderson KD, et al. Age-specific effects of hormone therapy use on overall mortality and ischemic heart disease mortality among women in the California Teachers Study. Menopause 2011;18(3):253-261.

Allison MA, Manson JE. Age, hormone therapy use, coronary heart disease, and mortality [editorial]. Menopause. 2011;18(3):243-245.

The initial findings of the WHI estrogen-progestin arm suggested that menopausal HT increases the risk of CHD. Since then, however, further analyses from the WHI and other HT trials, as well as reports from the observational Nurses’ Health Study, have suggested that the timing of initiation of HT determines its effect on cardiovascular health.

In this study from the California Teachers Study (CTS), investigators explored the effect of age at initiation of HT on cardiovascular and overall mortality. The CTS is a prospective study of more than 133,000 current and retired female teachers and administrators who returned an initial questionnaire in 1995 and 1996. Participants were then followed until late 2004, or death, whichever came first. More than 71,000 participants were eligible for analysis.

Current HT users were leaner, less likely to smoke, and more likely to exercise and consume alcohol than nonusers were. The analysis was adjusted for a variety of potential cardiovascular and other confounders.

Youngest HT users had the lowest risk of death

During follow-up, 18.3% of never-users of HT died, compared with 17.9% of former users. In contrast, 6.9% of women taking HT at the time of the baseline questionnaire died during follow-up.

Overall, current HT use was associated with a reduced risk of death from CHD (hazard ratio [HR], 0.84; 95% confidence interval, 0.74–0.95). This risk reduction was most notable (HR, 0.38) in the youngest HT users (36 to 59 years old). The risk of death from CHD gradually increased with the age of current HT users, reaching a hazard ratio of approximately 0.9 in current users who were 70 years and older. However, the CHD mortality hazard ratio did not reach or exceed the referent hazard ratio (1.0) assigned to never users of HT of any age.

The overall mortality rate was lowest for the youngest HT users (HR, 0.54) and approached 1.0 in the oldest current HT users.

The associations between overall and CHD mortality were similar among users of estrogen-only and estrogen-progestin HT.

Hormone therapy and dementia: Earlier use is better

Whitmer RA, Quesenberry CP, Zhou J, Yaffe K. Timing of hormone therapy and dementia: the critical window theory revisited. Ann Neurol. 2011;69(1):163–169.

Alzheimer’s disease is more common among women than men. In addition, caregivers to those who have dementia are more likely to be women. Therefore, it’s no surprise that women are especially concerned about their risk of dementia. Menopausal patients in my practice often ask whether use of HT might alter this risk.

Because vasomotor symptoms usually arise in late perimenopause or early menopause, women in observational studies (which reflect clinical practice) tend to begin HT when they are in their late 40s or early 50s. Overall, observational studies have suggested that HT is associated with a reduced risk of dementia. In contrast, the WHI clinical trial, in which the mean age of women who were randomized to HT or placebo was 63 years, found that the initiation of HT later in life increased the risk of dementia.

These observations led to the “critical window” theory regarding HT and dementia: Estrogen protects against dementia when it is taken by perimenopausal or early menopausal women, whereas it is not protective and may even accelerate cognitive decline when it is started many years after the onset of menopause.

In this recent study from the California Kaiser Permanente health maintenance organization, investigators assessed the long-term risk of dementia by timing of HT. From 1964 through 1973, menopausal “midlife” women who were 40 to 55 years old and free of dementia reported whether or not they used HT. Twenty-five to 30 years later, participants were reassessed for “late life” HT use.

Women who used HT in midlife only had the lowest prevalence of dementia, whereas those who used HT only in late life had the highest prevalence. Women who used HT at both time points had a prevalence of dementia similar to that of women who had never used HT.

Are SRIs an effective alternative to HT for hot flushes?

Freeman EW, Guthrie K, Caan B, et al. Efficacy of escitalopram for hot flashes in healthy menopausal women: a randomized controlled trial. JAMA. 2011;305(3):267–274.

Interest in nonhormonal management of menopausal vasomotor symptoms continues to run high, although only hormonal therapy has FDA approval for this indication. Many trials of serotonin reuptake inhibitors (SRIs) for the treatment of vasomotor symptoms have focused on breast cancer survivors, many of whom use anti-estrogen agents that increase the prevalence of these symptoms. In contrast, this well-conducted multicenter trial, funded by the National Institutes of Health, enrolled healthy, symptomatic, menopausal women.

In the trial, 205 perimenopausal or postmenopausal women 40 to 62 years old who had at least 28 bothersome or severe episodes of hot flushes and night sweats a week were randomized to 10 mg daily of the SRI escitalopram (Lexapro) or placebo for 8 weeks. Women who did not report a reduction in hot flushes and night sweats of at least 50% at 4 weeks, or a decrease in the severity of these symptoms, were increased to a dosage of 20 mg daily of escitalopram or placebo. The mean baseline frequency of vasomotor symptoms was 9.79.

Within 1 week, women taking the SRI experienced significantly greater improvement than those taking placebo. By 8 weeks, the daily frequency of vasomotor symptoms had diminished by 4.6 hot flushes among women taking the SRI, compared with 3.20 among women taking placebo (P < .01).

Overall, adverse effects were reported by approximately 58% of participants. The pattern of these side effects was similar in the active and placebo treatment arms. No adverse events serious enough to require withdrawal from the study were reported in either arm.

Patient satisfaction with treatment was 70% in the SRI group, compared with 43% among women taking placebo (P < .001).

Unopposed estrogen appears to protect against breast cancer

LaCroix AZ, Chlebowski RT, Manson JE, et al; WHI Investigators. Health outcomes after stopping conjugated equine estrogens among postmenopausal women with prior hysterectomy. A randomized controlled trial. JAMA. 2011;305(13):1305–1314.

The WHI continues to surprise with its findings almost a decade after publication of initial data. In this brand new report from the estrogen-alone arm, postmenopausal, hysterectomized women who were followed for a mean of 10.7 years experienced a reduced risk of breast cancer after a mean of 5.9 years of use of conjugated equine estrogens (CEE).

They experienced no increased or diminished risk of coronary heart disease (CHD), deep venous thrombosis, stroke, hip fracture, colorectal cancer, or total mortality after post-intervention follow-up.

Keep in mind that the women in this arm were instructed to discontinue the study medication at the time the intervention phase was halted because of an increased risk of stroke among CEE users. The elevated risk of stroke attenuated with the longer follow-up.

All ages experienced a reduced risk of breast cancer

Some subgroup analyses from the WHI have found differential effects of HT by age of the user, with younger women experiencing fewer risks and more benefits than those who are more than 10 years past the menopausal transition. In this analysis, all three age groups (50–59 years, 60–69 years, and 70–79 years) of women who used CEE had a reduced risk of breast cancer, compared with placebo users.

Other risks did appear to differ by age. For example, the overall hazard ratio for CHD was 0.59 among CEE users 50 to 59 years old, but it approached unity among the older women.

We want to hear from you! Tell us what you think.

Dr. Kaunitz receives grant or research support from Bayer, Agile, Noven, Teva, and Medical Diagnostic Laboratories, is a consultant to Bayer, Merck, and Teva, and owns stock in Becton Dickinson.

Among the developments of the past year in the care of menopausal women are:

• updated guidelines from the Institute of Medicine regarding vitamin D requirements—suggesting that fewer women are deficient in this nutrient than experts had believed
• new data from Europe on hormone therapy (HT) that highlight the safety of transdermal estrogen in comparison with oral administration
• a recent analysis from the Women’s Health Initiative (WHI), confirming a small elevated risk of breast cancer mortality with use of combination estrogen-progestin HT
• confirmation that age at initiation of HT determines its effect on cardiovascular health
• clarification of the association between HT and dementia
• new data demonstrating modest improvement in hot flushes when the serotonin reuptake inhibitor (SRI) escitalopram is used
• a brand new report from the WHI estrogen-alone arm that shows a protective effect against breast cancer.

The new data on HT suggest that we still have much to learn about its benefits and risks. We also are reaching an understanding that, for many young, symptomatic, menopausal patients, HT can represent a safe choice, with much depending on the timing and duration of therapy.

For more on how your colleagues are managing menopausal patients with and without hormone therapy, see “Is hormone therapy still a valid option? 12 ObGyns address this question,” on the facing page.

Menopausal women need less vitamin D than we thought

Institute of Medicine. Dietary reference intakes for calcium and vitamin D. Washington, DC: IOM; December 2010. http://www.iom.edu/~/media/Files/Report%20Files/2010/Dietary-Reference-Intakes-for-Calcium-and-Vitamin-D/Vitamin%20D%20and%20Calcium%202010%20Report%20Brief.pdf. Accessed March 24, 2011.

In the 2010 Update on Menopause, I summarized recent findings on vitamin D requirements, including recommendations that menopausal women should take at least 800 IU of vitamin D daily. I also described the prevailing expert opinion that many North American women are deficient in this nutrient.

What a difference a year can make! In late November, the Institute of Medicine (IOM) released a comprehensive report on vitamin D. Here are some of its conclusions:

• Vitamin D plays an important role in skeletal health but its role in other areas, including cardiovascular disease and cancer, is uncertain
• An intake of 600 IU of vitamin D daily is appropriate for girls and for women as old as 70 years; an in-take of 800 IU daily is appropriate for women older than 70 years
• A serum level of 25-hydroxy vitamin D of 20 ng/mL is consistent with adequate vitamin D status; this is lower than the threshold many have recommended
• With few exceptions, all people who live in North America—including those who have minimal or no exposure to sunlight—are receiving adequate calcium and vitamin D
• Ingestion of more than 4,000 IU of vitamin D daily can cause renal damage and injure other tissues.

The IOM report will likely prompt multivitamin manufacturers to increase the amount of vitamin D contained in their supplements to 600 IU daily. In addition, the report will probably discourage the common practice of checking serum 25-hydroxy vitamin D levels and prescribing a high dosage of vitamin D supplementation when the level is below 30 ng/mL.

Is transdermal estrogen safer than oral administration?

Canonico M, Fournier A, Carcaillon L, et al. Postmenopausal hormone therapy and risk of idiopathic venous thromboembolism: results from the E3N cohort study. Arterioscler Thromb Vasc Biol. 2010;30(2):340–345.

Renoux C, Dell’aniello S, Garbe E, Suissa S. Transdermal and oral hormone replacement therapy and the risk of stroke: a nested case-control study. BMJ. 2010;340:c2519. doi: 10.1136/bmj.c2519.

In the WHI, the combination of oral conjugated equine estrogen and medroxyprogesterone acetate more than doubled the risk of deep venous thrombosis and pulmonary embolism and modestly increased the risk of stroke, compared with nonuse.¹

A year after publication of the initial findings of the WHI estrogen-progestin arm, the Estrogen and THromboEmbolism Risk Study Group (ESTHER) case-control study from France provided evidence that transdermal estrogen does not increase the risk of venous thrombosis.² In France, many menopausal women use HT, and the transdermal route of administration is common.

In 2010, the E3N cohort study from France also assessed the risk of thrombosis associated with oral and transdermal HT. Investigators followed more than 80,000 postmenopausal women and found that, unlike oral HT, the transdermal route did not increase the risk of venous thrombosis.

More recent evidence also suggests a safety advantage for transdermal HT. The newest data come from the United Kingdom General Practice Research Database, which includes information on more than 870,000 women who were 50 to 70 years old from 1987 to 2006. Investigators identified more than 15,000 women who were given a diagnosis of stroke during this period and compared the use of HT in these women with that of almost 60,000 women in a control group. The risk of stroke associated with current use of transdermal HT was similar to the risk associated with nonuse of HT. Women who used a patch containing 0.05 mg of estradiol or less had a risk of stroke 19% lower than women who did not use HT.

In contrast, the risk of stroke in users of patches that contained a higher dosage of estradiol was almost twice the risk in nonusers of HT. Current users of oral HT had a risk of stroke 28% higher than that of nonusers of HT.

Estrogen-progestin HT raises the risk of death from breast cancer

Chlebowski RT, Anderson GL, Gass M, et al. Estrogen plus progestin and breast cancer incidence and mortality in postmenopausal women. JAMA. 2010;304(15):1684–1692.

Toh S, Hernandez-Diaz S, Logan R, Rossouw JE, Hernan MA. Coronary heart disease in postmenopausal recipients of estrogen plus progestin: does the increased risk ever disappear? Ann Intern Med. 2010;152(4):211–217.

In the estrogen-progestin arm of the WHI, initially published in 2002, the risk of invasive breast cancer was modestly elevated (hazard ratio [HR], 1.26) among women who had used HT longer than 5 years.³

In 2010, investigators reported on breast cancer mortality in WHI participants at a mean follow-up of 11 years. They found that combination HT users had breast cancer histology similar to that of nonusers. However, the tumors were more likely to be node-positive in combination HT users (23.7% vs 16.2%). In addition, breast cancer mortality was slightly higher among users of HT (2.6 vs 1.3 deaths in every 10,000 woman-years) (HR, 1.96; 95% confidence interval, 1.00–4.04).

Earlier observational studies had suggested that the death rate from breast cancer is lower in users of combination HT than in nonusers. Consistent with the UK Million Women Study, however, a 2010 report from the WHI found a higher mortality rate among women who have used HT.⁴

These new WHI findings reinforce the importance of assessing whether micronized progesterone combined with estrogen might lower the risk of death from breast cancer—a possibility suggested by findings of the French E3N cohort study.⁵

In addition, given the possibility that HT may be cardioprotective when it is initiated within 10 years after the onset of menopause, a WHI report that addresses long-term all-cause mortality would allow us to better counsel our menopausal patients who are trying to decide whether to start or continue HT. See, for example, the data from the California Teachers Study (below) and the estrogen-alone arm of the WHI (page 46).

Age at initiation of HT determines its effect on CHD

Stram DO, Liu Y, Henderson KD, et al. Age-specific effects of hormone therapy use on overall mortality and ischemic heart disease mortality among women in the California Teachers Study. Menopause 2011;18(3):253-261.

Allison MA, Manson JE. Age, hormone therapy use, coronary heart disease, and mortality [editorial]. Menopause. 2011;18(3):243-245.

The initial findings of the WHI estrogen-progestin arm suggested that menopausal HT increases the risk of CHD. Since then, however, further analyses from the WHI and other HT trials, as well as reports from the observational Nurses’ Health Study, have suggested that the timing of initiation of HT determines its effect on cardiovascular health.

In this study from the California Teachers Study (CTS), investigators explored the effect of age at initiation of HT on cardiovascular and overall mortality. The CTS is a prospective study of more than 133,000 current and retired female teachers and administrators who returned an initial questionnaire in 1995 and 1996. Participants were then followed until late 2004, or death, whichever came first. More than 71,000 participants were eligible for analysis.

Current HT users were leaner, less likely to smoke, and more likely to exercise and consume alcohol than nonusers were. The analysis was adjusted for a variety of potential cardiovascular and other confounders.

Youngest HT users had the lowest risk of death

During follow-up, 18.3% of never-users of HT died, compared with 17.9% of former users. In contrast, 6.9% of women taking HT at the time of the baseline questionnaire died during follow-up.

Overall, current HT use was associated with a reduced risk of death from CHD (hazard ratio [HR], 0.84; 95% confidence interval, 0.74–0.95). This risk reduction was most notable (HR, 0.38) in the youngest HT users (36 to 59 years old). The risk of death from CHD gradually increased with the age of current HT users, reaching a hazard ratio of approximately 0.9 in current users who were 70 years and older. However, the CHD mortality hazard ratio did not reach or exceed the referent hazard ratio (1.0) assigned to never users of HT of any age.

The overall mortality rate was lowest for the youngest HT users (HR, 0.54) and approached 1.0 in the oldest current HT users.

The associations between overall and CHD mortality were similar among users of estrogen-only and estrogen-progestin HT.

Hormone therapy and dementia: Earlier use is better

Whitmer RA, Quesenberry CP, Zhou J, Yaffe K. Timing of hormone therapy and dementia: the critical window theory revisited. Ann Neurol. 2011;69(1):163–169.

Alzheimer’s disease is more common among women than men. In addition, caregivers to those who have dementia are more likely to be women. Therefore, it’s no surprise that women are especially concerned about their risk of dementia. Menopausal patients in my practice often ask whether use of HT might alter this risk.

Because vasomotor symptoms usually arise in late perimenopause or early menopause, women in observational studies (which reflect clinical practice) tend to begin HT when they are in their late 40s or early 50s. Overall, observational studies have suggested that HT is associated with a reduced risk of dementia. In contrast, the WHI clinical trial, in which the mean age of women who were randomized to HT or placebo was 63 years, found that the initiation of HT later in life increased the risk of dementia.

These observations led to the “critical window” theory regarding HT and dementia: Estrogen protects against dementia when it is taken by perimenopausal or early menopausal women, whereas it is not protective and may even accelerate cognitive decline when it is started many years after the onset of menopause.

In this recent study from the California Kaiser Permanente health maintenance organization, investigators assessed the long-term risk of dementia by timing of HT. From 1964 through 1973, menopausal “midlife” women who were 40 to 55 years old and free of dementia reported whether or not they used HT. Twenty-five to 30 years later, participants were reassessed for “late life” HT use.

Women who used HT in midlife only had the lowest prevalence of dementia, whereas those who used HT only in late life had the highest prevalence. Women who used HT at both time points had a prevalence of dementia similar to that of women who had never used HT.

Are SRIs an effective alternative to HT for hot flushes?

Freeman EW, Guthrie K, Caan B, et al. Efficacy of escitalopram for hot flashes in healthy menopausal women: a randomized controlled trial. JAMA. 2011;305(3):267–274.

Interest in nonhormonal management of menopausal vasomotor symptoms continues to run high, although only hormonal therapy has FDA approval for this indication. Many trials of serotonin reuptake inhibitors (SRIs) for the treatment of vasomotor symptoms have focused on breast cancer survivors, many of whom use anti-estrogen agents that increase the prevalence of these symptoms. In contrast, this well-conducted multicenter trial, funded by the National Institutes of Health, enrolled healthy, symptomatic, menopausal women.

In the trial, 205 perimenopausal or postmenopausal women 40 to 62 years old who had at least 28 bothersome or severe episodes of hot flushes and night sweats a week were randomized to 10 mg daily of the SRI escitalopram (Lexapro) or placebo for 8 weeks. Women who did not report a reduction in hot flushes and night sweats of at least 50% at 4 weeks, or a decrease in the severity of these symptoms, were increased to a dosage of 20 mg daily of escitalopram or placebo. The mean baseline frequency of vasomotor symptoms was 9.79.

Within 1 week, women taking the SRI experienced significantly greater improvement than those taking placebo. By 8 weeks, the daily frequency of vasomotor symptoms had diminished by 4.6 hot flushes among women taking the SRI, compared with 3.20 among women taking placebo (P < .01).

Overall, adverse effects were reported by approximately 58% of participants. The pattern of these side effects was similar in the active and placebo treatment arms. No adverse events serious enough to require withdrawal from the study were reported in either arm.

Patient satisfaction with treatment was 70% in the SRI group, compared with 43% among women taking placebo (P < .001).

Unopposed estrogen appears to protect against breast cancer

LaCroix AZ, Chlebowski RT, Manson JE, et al; WHI Investigators. Health outcomes after stopping conjugated equine estrogens among postmenopausal women with prior hysterectomy. A randomized controlled trial. JAMA. 2011;305(13):1305–1314.

The WHI continues to surprise with its findings almost a decade after publication of initial data. In this brand new report from the estrogen-alone arm, postmenopausal, hysterectomized women who were followed for a mean of 10.7 years experienced a reduced risk of breast cancer after a mean of 5.9 years of use of conjugated equine estrogens (CEE).

They experienced no increased or diminished risk of coronary heart disease (CHD), deep venous thrombosis, stroke, hip fracture, colorectal cancer, or total mortality after post-intervention follow-up.

Keep in mind that the women in this arm were instructed to discontinue the study medication at the time the intervention phase was halted because of an increased risk of stroke among CEE users. The elevated risk of stroke attenuated with the longer follow-up.

All ages experienced a reduced risk of breast cancer

Some subgroup analyses from the WHI have found differential effects of HT by age of the user, with younger women experiencing fewer risks and more benefits than those who are more than 10 years past the menopausal transition. In this analysis, all three age groups (50–59 years, 60–69 years, and 70–79 years) of women who used CEE had a reduced risk of breast cancer, compared with placebo users.

Other risks did appear to differ by age. For example, the overall hazard ratio for CHD was 0.59 among CEE users 50 to 59 years old, but it approached unity among the older women.

We want to hear from you! Tell us what you think.

Dr. Kaunitz receives grant or research support from Bayer, Agile, Noven, Teva, and Medical Diagnostic Laboratories, is a consultant to Bayer, Merck, and Teva, and owns stock in Becton Dickinson.

Among the developments of the past year in the care of menopausal women are:

• updated guidelines from the Institute of Medicine regarding vitamin D requirements—suggesting that fewer women are deficient in this nutrient than experts had believed
• new data from Europe on hormone therapy (HT) that highlight the safety of transdermal estrogen in comparison with oral administration
• a recent analysis from the Women’s Health Initiative (WHI), confirming a small elevated risk of breast cancer mortality with use of combination estrogen-progestin HT
• confirmation that age at initiation of HT determines its effect on cardiovascular health
• clarification of the association between HT and dementia
• new data demonstrating modest improvement in hot flushes when the serotonin reuptake inhibitor (SRI) escitalopram is used
• a brand new report from the WHI estrogen-alone arm that shows a protective effect against breast cancer.

The new data on HT suggest that we still have much to learn about its benefits and risks. We also are reaching an understanding that, for many young, symptomatic, menopausal patients, HT can represent a safe choice, with much depending on the timing and duration of therapy.

For more on how your colleagues are managing menopausal patients with and without hormone therapy, see “Is hormone therapy still a valid option? 12 ObGyns address this question,” on the facing page.

Menopausal women need less vitamin D than we thought

Institute of Medicine. Dietary reference intakes for calcium and vitamin D. Washington, DC: IOM; December 2010. http://www.iom.edu/~/media/Files/Report%20Files/2010/Dietary-Reference-Intakes-for-Calcium-and-Vitamin-D/Vitamin%20D%20and%20Calcium%202010%20Report%20Brief.pdf. Accessed March 24, 2011.

In the 2010 Update on Menopause, I summarized recent findings on vitamin D requirements, including recommendations that menopausal women should take at least 800 IU of vitamin D daily. I also described the prevailing expert opinion that many North American women are deficient in this nutrient.

What a difference a year can make! In late November, the Institute of Medicine (IOM) released a comprehensive report on vitamin D. Here are some of its conclusions:

• Vitamin D plays an important role in skeletal health but its role in other areas, including cardiovascular disease and cancer, is uncertain
• An intake of 600 IU of vitamin D daily is appropriate for girls and for women as old as 70 years; an in-take of 800 IU daily is appropriate for women older than 70 years
• A serum level of 25-hydroxy vitamin D of 20 ng/mL is consistent with adequate vitamin D status; this is lower than the threshold many have recommended
• With few exceptions, all people who live in North America—including those who have minimal or no exposure to sunlight—are receiving adequate calcium and vitamin D
• Ingestion of more than 4,000 IU of vitamin D daily can cause renal damage and injure other tissues.

The IOM report will likely prompt multivitamin manufacturers to increase the amount of vitamin D contained in their supplements to 600 IU daily. In addition, the report will probably discourage the common practice of checking serum 25-hydroxy vitamin D levels and prescribing a high dosage of vitamin D supplementation when the level is below 30 ng/mL.

Is transdermal estrogen safer than oral administration?

Canonico M, Fournier A, Carcaillon L, et al. Postmenopausal hormone therapy and risk of idiopathic venous thromboembolism: results from the E3N cohort study. Arterioscler Thromb Vasc Biol. 2010;30(2):340–345.

Renoux C, Dell’aniello S, Garbe E, Suissa S. Transdermal and oral hormone replacement therapy and the risk of stroke: a nested case-control study. BMJ. 2010;340:c2519. doi: 10.1136/bmj.c2519.

In the WHI, the combination of oral conjugated equine estrogen and medroxyprogesterone acetate more than doubled the risk of deep venous thrombosis and pulmonary embolism and modestly increased the risk of stroke, compared with nonuse.¹

A year after publication of the initial findings of the WHI estrogen-progestin arm, the Estrogen and THromboEmbolism Risk Study Group (ESTHER) case-control study from France provided evidence that transdermal estrogen does not increase the risk of venous thrombosis.² In France, many menopausal women use HT, and the transdermal route of administration is common.

In 2010, the E3N cohort study from France also assessed the risk of thrombosis associated with oral and transdermal HT. Investigators followed more than 80,000 postmenopausal women and found that, unlike oral HT, the transdermal route did not increase the risk of venous thrombosis.

More recent evidence also suggests a safety advantage for transdermal HT. The newest data come from the United Kingdom General Practice Research Database, which includes information on more than 870,000 women who were 50 to 70 years old from 1987 to 2006. Investigators identified more than 15,000 women who were given a diagnosis of stroke during this period and compared the use of HT in these women with that of almost 60,000 women in a control group. The risk of stroke associated with current use of transdermal HT was similar to the risk associated with nonuse of HT. Women who used a patch containing 0.05 mg of estradiol or less had a risk of stroke 19% lower than women who did not use HT.

In contrast, the risk of stroke in users of patches that contained a higher dosage of estradiol was almost twice the risk in nonusers of HT. Current users of oral HT had a risk of stroke 28% higher than that of nonusers of HT.

Estrogen-progestin HT raises the risk of death from breast cancer

Chlebowski RT, Anderson GL, Gass M, et al. Estrogen plus progestin and breast cancer incidence and mortality in postmenopausal women. JAMA. 2010;304(15):1684–1692.

Toh S, Hernandez-Diaz S, Logan R, Rossouw JE, Hernan MA. Coronary heart disease in postmenopausal recipients of estrogen plus progestin: does the increased risk ever disappear? Ann Intern Med. 2010;152(4):211–217.

In the estrogen-progestin arm of the WHI, initially published in 2002, the risk of invasive breast cancer was modestly elevated (hazard ratio [HR], 1.26) among women who had used HT longer than 5 years.³

In 2010, investigators reported on breast cancer mortality in WHI participants at a mean follow-up of 11 years. They found that combination HT users had breast cancer histology similar to that of nonusers. However, the tumors were more likely to be node-positive in combination HT users (23.7% vs 16.2%). In addition, breast cancer mortality was slightly higher among users of HT (2.6 vs 1.3 deaths in every 10,000 woman-years) (HR, 1.96; 95% confidence interval, 1.00–4.04).

Earlier observational studies had suggested that the death rate from breast cancer is lower in users of combination HT than in nonusers. Consistent with the UK Million Women Study, however, a 2010 report from the WHI found a higher mortality rate among women who have used HT.⁴

These new WHI findings reinforce the importance of assessing whether micronized progesterone combined with estrogen might lower the risk of death from breast cancer—a possibility suggested by findings of the French E3N cohort study.⁵

In addition, given the possibility that HT may be cardioprotective when it is initiated within 10 years after the onset of menopause, a WHI report that addresses long-term all-cause mortality would allow us to better counsel our menopausal patients who are trying to decide whether to start or continue HT. See, for example, the data from the California Teachers Study (below) and the estrogen-alone arm of the WHI (page 46).

Age at initiation of HT determines its effect on CHD

Stram DO, Liu Y, Henderson KD, et al. Age-specific effects of hormone therapy use on overall mortality and ischemic heart disease mortality among women in the California Teachers Study. Menopause 2011;18(3):253-261.

Allison MA, Manson JE. Age, hormone therapy use, coronary heart disease, and mortality [editorial]. Menopause. 2011;18(3):243-245.

The initial findings of the WHI estrogen-progestin arm suggested that menopausal HT increases the risk of CHD. Since then, however, further analyses from the WHI and other HT trials, as well as reports from the observational Nurses’ Health Study, have suggested that the timing of initiation of HT determines its effect on cardiovascular health.

In this study from the California Teachers Study (CTS), investigators explored the effect of age at initiation of HT on cardiovascular and overall mortality. The CTS is a prospective study of more than 133,000 current and retired female teachers and administrators who returned an initial questionnaire in 1995 and 1996. Participants were then followed until late 2004, or death, whichever came first. More than 71,000 participants were eligible for analysis.

Current HT users were leaner, less likely to smoke, and more likely to exercise and consume alcohol than nonusers were. The analysis was adjusted for a variety of potential cardiovascular and other confounders.

Youngest HT users had the lowest risk of death

During follow-up, 18.3% of never-users of HT died, compared with 17.9% of former users. In contrast, 6.9% of women taking HT at the time of the baseline questionnaire died during follow-up.

Overall, current HT use was associated with a reduced risk of death from CHD (hazard ratio [HR], 0.84; 95% confidence interval, 0.74–0.95). This risk reduction was most notable (HR, 0.38) in the youngest HT users (36 to 59 years old). The risk of death from CHD gradually increased with the age of current HT users, reaching a hazard ratio of approximately 0.9 in current users who were 70 years and older. However, the CHD mortality hazard ratio did not reach or exceed the referent hazard ratio (1.0) assigned to never users of HT of any age.

The overall mortality rate was lowest for the youngest HT users (HR, 0.54) and approached 1.0 in the oldest current HT users.

The associations between overall and CHD mortality were similar among users of estrogen-only and estrogen-progestin HT.

Hormone therapy and dementia: Earlier use is better

Whitmer RA, Quesenberry CP, Zhou J, Yaffe K. Timing of hormone therapy and dementia: the critical window theory revisited. Ann Neurol. 2011;69(1):163–169.

Alzheimer’s disease is more common among women than men. In addition, caregivers to those who have dementia are more likely to be women. Therefore, it’s no surprise that women are especially concerned about their risk of dementia. Menopausal patients in my practice often ask whether use of HT might alter this risk.

Because vasomotor symptoms usually arise in late perimenopause or early menopause, women in observational studies (which reflect clinical practice) tend to begin HT when they are in their late 40s or early 50s. Overall, observational studies have suggested that HT is associated with a reduced risk of dementia. In contrast, the WHI clinical trial, in which the mean age of women who were randomized to HT or placebo was 63 years, found that the initiation of HT later in life increased the risk of dementia.

These observations led to the “critical window” theory regarding HT and dementia: Estrogen protects against dementia when it is taken by perimenopausal or early menopausal women, whereas it is not protective and may even accelerate cognitive decline when it is started many years after the onset of menopause.

In this recent study from the California Kaiser Permanente health maintenance organization, investigators assessed the long-term risk of dementia by timing of HT. From 1964 through 1973, menopausal “midlife” women who were 40 to 55 years old and free of dementia reported whether or not they used HT. Twenty-five to 30 years later, participants were reassessed for “late life” HT use.

Women who used HT in midlife only had the lowest prevalence of dementia, whereas those who used HT only in late life had the highest prevalence. Women who used HT at both time points had a prevalence of dementia similar to that of women who had never used HT.

Are SRIs an effective alternative to HT for hot flushes?

Freeman EW, Guthrie K, Caan B, et al. Efficacy of escitalopram for hot flashes in healthy menopausal women: a randomized controlled trial. JAMA. 2011;305(3):267–274.

Interest in nonhormonal management of menopausal vasomotor symptoms continues to run high, although only hormonal therapy has FDA approval for this indication. Many trials of serotonin reuptake inhibitors (SRIs) for the treatment of vasomotor symptoms have focused on breast cancer survivors, many of whom use anti-estrogen agents that increase the prevalence of these symptoms. In contrast, this well-conducted multicenter trial, funded by the National Institutes of Health, enrolled healthy, symptomatic, menopausal women.

In the trial, 205 perimenopausal or postmenopausal women 40 to 62 years old who had at least 28 bothersome or severe episodes of hot flushes and night sweats a week were randomized to 10 mg daily of the SRI escitalopram (Lexapro) or placebo for 8 weeks. Women who did not report a reduction in hot flushes and night sweats of at least 50% at 4 weeks, or a decrease in the severity of these symptoms, were increased to a dosage of 20 mg daily of escitalopram or placebo. The mean baseline frequency of vasomotor symptoms was 9.79.

Within 1 week, women taking the SRI experienced significantly greater improvement than those taking placebo. By 8 weeks, the daily frequency of vasomotor symptoms had diminished by 4.6 hot flushes among women taking the SRI, compared with 3.20 among women taking placebo (P < .01).

Overall, adverse effects were reported by approximately 58% of participants. The pattern of these side effects was similar in the active and placebo treatment arms. No adverse events serious enough to require withdrawal from the study were reported in either arm.

Patient satisfaction with treatment was 70% in the SRI group, compared with 43% among women taking placebo (P < .001).

Unopposed estrogen appears to protect against breast cancer

LaCroix AZ, Chlebowski RT, Manson JE, et al; WHI Investigators. Health outcomes after stopping conjugated equine estrogens among postmenopausal women with prior hysterectomy. A randomized controlled trial. JAMA. 2011;305(13):1305–1314.

The WHI continues to surprise with its findings almost a decade after publication of initial data. In this brand new report from the estrogen-alone arm, postmenopausal, hysterectomized women who were followed for a mean of 10.7 years experienced a reduced risk of breast cancer after a mean of 5.9 years of use of conjugated equine estrogens (CEE).

They experienced no increased or diminished risk of coronary heart disease (CHD), deep venous thrombosis, stroke, hip fracture, colorectal cancer, or total mortality after post-intervention follow-up.

Keep in mind that the women in this arm were instructed to discontinue the study medication at the time the intervention phase was halted because of an increased risk of stroke among CEE users. The elevated risk of stroke attenuated with the longer follow-up.

All ages experienced a reduced risk of breast cancer

Some subgroup analyses from the WHI have found differential effects of HT by age of the user, with younger women experiencing fewer risks and more benefits than those who are more than 10 years past the menopausal transition. In this analysis, all three age groups (50–59 years, 60–69 years, and 70–79 years) of women who used CEE had a reduced risk of breast cancer, compared with placebo users.

Other risks did appear to differ by age. For example, the overall hazard ratio for CHD was 0.59 among CEE users 50 to 59 years old, but it approached unity among the older women.

We want to hear from you! Tell us what you think.

CAS: Depressive disorder, anticipating a pregnancy

Your patient Megan—well-educated, 29 years old, G0P0—has come to you to discuss her antidepressant (paroxetine [Paxil]) because she is planning her first pregnancy.

Megan has a history of recurrent major depressive disorder (MDD), which is in remission (see “What is MDD?”).

How will you begin the conversation with this patient about keeping MDD in remission during her pregnancy and ensuring the safety of her fetus?

There is a 20% to 25% lifetime prevalence of depression in women; the disorder peaks during childbearing years, however.¹ As of 2003, 13% of pregnant women had taken an antidepressant at some time during their pregnancy, a percentage that has doubled since it was assessed in 1999.²

You are faced with several quandaries in deciding whether to recommend that your patient continue, or discontinue, antidepressant therapy during pregnancy:

• As many as 68% of women who terminate antidepressant treatment before or during pregnancy relapse.
• Even 26% of women who continue antidepressant during pregnancy relapse—requiring a dosage adjustment or change in treatment.³
• Yet the possibly elevated cortisol levels of severe, untreated depression may harm the placenta and fetus.^4,5

So, what do you need to know to assess the risks and benefits of “Megan” stopping, or continuing, paroxetine during her anticipated pregnancy? And what are the risks to Megan’s fetus of treating, or not treating, her depression with a serotonin reuptake inhibitor (SRI*)?

Gauging the risks of depression in pregnancy

In any given patient, her history and family history of depression are key to determining the likelihood that she will suffer ongoing or recurrent depression.

CASE continued Repeated treated episodes plus a family history

In obtaining Megan’s history, you learn that she has had three prior episodes of depression, all of which were successfully treated with paroxetine. Megan has been stable on paroxetine for 3 years.

Notably, the second episode of depression was initially treated with a 16-week trial of psychotherapy alone; when depressive symptoms did not remit, paroxetine was added. That episode was considered severe because it included pervasive thoughts of suicide.

You also learn that Megan’s mother suffered from postpartum depression and that her father and paternal grandmother were treated for depression.

Known risk factors for depression during pregnancy include: maternal anxiety; prior diagnosis of depression during pregnancy; history of postpartum anxiety or depression; prior diagnosis of either anxiety or depressive disorder; significant life stress (e.g., divorce, death of a loved one); degree of social support—particularly, intimate social support; “intendedness” of pregnancy; domestic violence; and insurance status.⁶

You review with Megan her risk factors for depression during pregnancy, namely: three prior episodes of MDD and a strong family history of mood disorder. Her MDD is considered “severe” because she has a history of suicidality. You tell Megan that, given these factors, she is at high risk of a recurrence of her depressive illness during pregnancy.

Megan asks: “Would getting depressed during pregnancy hurt the baby?”

Depression during pregnancy affects both infant and maternal well-being, although studies are in conflict about the extent of that morbidity. Multiple areas of potential risk to mother and infant have been studied, including the effect of depression on:

• maternal well-being
• growth of the infant
• spontaneous abortion
• preterm delivery
• neonatal physiologic and neurobehavioral measures
• long-term considerations for the developing infant and child.

Within these categories of risk, a diagnosis of depression during pregnancy has been associated (in some but not all studies) with a higher risk, or rate, of:

• postpartum depression
• preterm birth
• lower maternal weight gain
• maternal tobacco, alcohol, and other substance use
• lower infant gestational age at birth
• small-for-gestational age infant birth.^7-10

In terms of long-term impact on offspring, studies differ in their estimation of risk; however, children exposed to untreated, maternal depression at 18 weeks’ and 32 weeks’ gestation did show a greater degree of developmental delay at 18 months than children who were born to a mother who was not depressed during pregnancy.¹¹

You discuss these risks with Megan. She asks: “What treatment do you recommend for me?” You turn to the 2009 guidelines published jointly by the American Psychiatric Association (APA) and ACOG.

These guidelines recommend that you consider 1) the severity of her current depression, 2) her history of depression severity, and 3) her preference for treatment.¹² For mild depression during pregnancy, when there is no history of severe depression, or for a history of depression that responded well to psychotherapy in the past, a trial of psychotherapy without medications is recommended.

But Megan’s history of depression falls into the “severe” category, and a prior episode of depression did not respond well to psychotherapy. Your recommendation to her, therefore, is that she should continue taking an antidepressant—unless she feels strongly that she should discontinue it.

Risks of SRIs in pregnancy

Megan considers what you’ve discussed about her high risk of developing recurrent depression during pregnancy. She decides that she wants to continue taking her antidepressant during pregnancy, but she has concerns—based on what she has been reading on the Internet.

Megan hands you a detailed printout downloaded from a Web site unfamiliar to you and asks about risks to the baby of such medications as paroxetine.

What should you tell Megan about SRIs in pregnancy—paroxetine, specifically?

You preface your remarks to her by noting that the data physicians work with are imperfect—because randomized, controlled clinical trials pose an ethical dilemma as a method of study in pregnant women. You then discuss with her current scientific understanding of potential risks to her fetus.

The difference in the rates of structural malformation among SRI-exposed and SRI-unexposed groups has been studied; most studies have found no increased rate of major or specific cardiac malformations.¹² However, first-trimester paroxetine appeared, in some studies, to be associated with an increased rate of cardiac malformations. That led to a category-“D” pregnancy classification in 2005 and an FDA “Public Health Advisory.”

Other large cohort studies have not uncovered such an association. It has been hypothesized that the methodology of data collection may have influenced this finding.¹³

Other malformations have been implicated in some studies but not others, and have included associations between specific SRIs and cardiac ventricular outflow defects, craniosynostosis, and omphalocele. The absolute risk of these defects remains extremely low, however, and close to the background rate seen in the general population.¹⁴

Megan asks: “With that risk-category ‘D’ for paroxetine, do you recommend I continue taking it or should I switch to another medication while I’m pregnant?”

You review again with Megan that, although some studies have linked first-trimester paroxetine to an increased risk of cardiac malformation, that finding has not been replicated in several large cohort studies. You explain that, if she had a history of recurrent depression that had failed to respond to many antidepressants and only paroxetine worked, an attempt at switching the SRI would not be recommended because of the potential for relapse.

Megan tells you that she would feel safer not taking a category-“D” drug. You agree and propose a judicious approach: Because she has come to see you before she became pregnant, with enough time to complete a slow crossover to an alternative SRI, and because she has not had any earlier trials of other SRIs, a slow taper of paroxetine, coupled with a crossover to an alternative SRI, is a reasonable option—with the caution that substitution always carries a risk of relapse.

Problems in newborns

Megan considers the risks you’ve discussed so far. She remembers a recent article in a magazine for pregnant women that described severe “respiratory” and “withdrawal” symptoms in infants who were born to mothers taking an SRI antidepressant. She wonders if she should consider discontinuing her SRI in the third trimester to try to mitigate those risks.

Megan is asking you about an SRI exposure risk that has been fairly consistent across studies, called neonatal abstinence syndrome (NAS) or poor neonatal adaptation.

NAS is a cluster of symptoms that occurs in 15% to 30% of newborns who have been exposed to an SRI during the third trimester of pregnancy.¹⁵ Signs include irritability, weak cry, tachypnea, temperature instability, and hypoglycemia—all of which are transient, peak during the first 48 hours after delivery, and resolve in less than 2 weeks.

Multiple hypotheses have been put forward to account for NAS, including the possibilities that it reflects a withdrawal syndrome, pharmacotoxicity, or an underlying gene–SRI interaction. The physiology behind NAS remains unknown, however.¹²

Megan next asks you about persistent pulmonary hypertension of the newborn (PPHN). You explain that PPHN is of recent concern in women who have been taking an SRI in the latter half of their pregnancy.

The rate of PPHN in the general population is 0.5 to 2 newborns for every 1,000. Associated mortality is approximately 10% to 20%.^16-18 This rate is thought to rise to approximately 6 of every 1,000 newborns among those who have been exposed to an SRI in utero—with some evidence of increased risk conferred through SRI exposure during later pregnancy (studies define this as the second half of the pregnancy).¹⁵ Although the relative risk of PPHN is increased threefold to sixfold when an SRI is used in pregnancy, absolute risk remains extremely low.

Concerns have been raised over research methodology in the few studies that have looked into SRI exposure and PPHN. Not all such studies found a change in relative risk or absolute risk of PPHN in SRI- exposed infants, compared to what was found in non-SRI–exposed infants.^15,19,20

Megan presses you, however, with the understandable question of whether she should taper her SRI during the last trimester (which the Web site she has found recommends). With the above information in mind, you explain that, given current understanding of the low absolute risk of PPHN, and given her illness history and severity of prior depression, you would not recommend that she taper the antidepressant in the third trimester.

Furthermore, the same counsel applies in regard to NAS: Given the risk of psychiatric morbidity caused by discontinuing an SRI during the third trimester, you do not recommend that she taper an SRI during that period to avoid NAS.

You explain that, instead, physicians now counsel women who take an SRI about the signs of NAS so that they can be prepared if they observe any of them in their infant.

Megan has one more question: “Will I be able to breastfeed while I’m taking an antidepressant?”

Postpartum issues to consider

Given the inherent difficulties and risks of relapse associated with a crossover to an alternative antidepressant postpartum, it makes sense, when possible, for a woman to take an antidepressant during pregnancy that can safely be continued while she is breastfeeding.

You tell Megan that, even though the quality of the data in this area is also thin, SRIs that have a low maternal serum profile are considered safest in breastfeeding.

To date, two SRIs—sertraline and paroxetine—have not been detectable in the breast milk of women taking either of them.²¹

CASE Appointment concluded, overflowing with information,
advice, and optimism

Megan says that, taking into account all that you and she have talked about, and even though she wants to return with her husband, she would like to switch to sertraline before she becomes pregnant—while she gauges its effectiveness at keeping her disorder in remission.

A good outcome requires you to prevail over obstacles

Because a diagnosis of depression spans a continuum of severity and, often, is not perceived as an acutely life-threatening illness, evaluating the risks and benefits of treatment is a murky undertaking.

Our role as physicians is to, first, educate ourselves and our patients about these variables and, second, support our patients in the decisions that they make. Physicians who care for pregnant women must be aware of the benefits and limitations of treatments as reported in the most current literature if they are going to assist women with decisions about treatment in the best possible way.

Social stigma. There remains the impact of stigma. Depressive and anxiety disorders are often perceived to be either under the control of an affected person’s “free will” or not as serious as other forms of “medical” disease. Consequently, the role that cultural and social pressures play in the risk–benefit analysis conducted by pregnant women and their physicians can’t be discounted.

Customized decision-making. As more data emerge about the treatment of depression in pregnancy, it has become clear: Treatment algorithms meant to simplify our decisions must always be individualized and extended into the postpartum period.

Treatment selection. Management of mild depression during pregnancy does not always require medication. Multiple variables—the list is long, and includes a patient’s psychiatric history, family psychiatric history, response to prior treatment, severity of depression, severity of prior depression, degree of social support, and personal desires—must be considered in determining what treatment is appropriate before, during, and after a pregnancy.

For a woman who suffers mild or moderate depression, with few antenatal depression risk factors, a trial of psychotherapy is recommended as first-line treatment. For a woman suffering from severe depression, or one who has a history of severe depression that has not responded well to psychotherapy alone, continuation or initiation of an SRI antidepressant is the current recommendation.

We want to hear from you! Tell us what you think.

CAS: Depressive disorder, anticipating a pregnancy

Your patient Megan—well-educated, 29 years old, G0P0—has come to you to discuss her antidepressant (paroxetine [Paxil]) because she is planning her first pregnancy.

Megan has a history of recurrent major depressive disorder (MDD), which is in remission (see “What is MDD?”).

How will you begin the conversation with this patient about keeping MDD in remission during her pregnancy and ensuring the safety of her fetus?

There is a 20% to 25% lifetime prevalence of depression in women; the disorder peaks during childbearing years, however.¹ As of 2003, 13% of pregnant women had taken an antidepressant at some time during their pregnancy, a percentage that has doubled since it was assessed in 1999.²

You are faced with several quandaries in deciding whether to recommend that your patient continue, or discontinue, antidepressant therapy during pregnancy:

• As many as 68% of women who terminate antidepressant treatment before or during pregnancy relapse.
• Even 26% of women who continue antidepressant during pregnancy relapse—requiring a dosage adjustment or change in treatment.³
• Yet the possibly elevated cortisol levels of severe, untreated depression may harm the placenta and fetus.^4,5

So, what do you need to know to assess the risks and benefits of “Megan” stopping, or continuing, paroxetine during her anticipated pregnancy? And what are the risks to Megan’s fetus of treating, or not treating, her depression with a serotonin reuptake inhibitor (SRI*)?

Gauging the risks of depression in pregnancy

In any given patient, her history and family history of depression are key to determining the likelihood that she will suffer ongoing or recurrent depression.

CASE continued Repeated treated episodes plus a family history

In obtaining Megan’s history, you learn that she has had three prior episodes of depression, all of which were successfully treated with paroxetine. Megan has been stable on paroxetine for 3 years.

Notably, the second episode of depression was initially treated with a 16-week trial of psychotherapy alone; when depressive symptoms did not remit, paroxetine was added. That episode was considered severe because it included pervasive thoughts of suicide.

You also learn that Megan’s mother suffered from postpartum depression and that her father and paternal grandmother were treated for depression.

Known risk factors for depression during pregnancy include: maternal anxiety; prior diagnosis of depression during pregnancy; history of postpartum anxiety or depression; prior diagnosis of either anxiety or depressive disorder; significant life stress (e.g., divorce, death of a loved one); degree of social support—particularly, intimate social support; “intendedness” of pregnancy; domestic violence; and insurance status.⁶

You review with Megan her risk factors for depression during pregnancy, namely: three prior episodes of MDD and a strong family history of mood disorder. Her MDD is considered “severe” because she has a history of suicidality. You tell Megan that, given these factors, she is at high risk of a recurrence of her depressive illness during pregnancy.

Megan asks: “Would getting depressed during pregnancy hurt the baby?”

Depression during pregnancy affects both infant and maternal well-being, although studies are in conflict about the extent of that morbidity. Multiple areas of potential risk to mother and infant have been studied, including the effect of depression on:

• maternal well-being
• growth of the infant
• spontaneous abortion
• preterm delivery
• neonatal physiologic and neurobehavioral measures
• long-term considerations for the developing infant and child.

Within these categories of risk, a diagnosis of depression during pregnancy has been associated (in some but not all studies) with a higher risk, or rate, of:

• postpartum depression
• preterm birth
• lower maternal weight gain
• maternal tobacco, alcohol, and other substance use
• lower infant gestational age at birth
• small-for-gestational age infant birth.^7-10

In terms of long-term impact on offspring, studies differ in their estimation of risk; however, children exposed to untreated, maternal depression at 18 weeks’ and 32 weeks’ gestation did show a greater degree of developmental delay at 18 months than children who were born to a mother who was not depressed during pregnancy.¹¹

You discuss these risks with Megan. She asks: “What treatment do you recommend for me?” You turn to the 2009 guidelines published jointly by the American Psychiatric Association (APA) and ACOG.

These guidelines recommend that you consider 1) the severity of her current depression, 2) her history of depression severity, and 3) her preference for treatment.¹² For mild depression during pregnancy, when there is no history of severe depression, or for a history of depression that responded well to psychotherapy in the past, a trial of psychotherapy without medications is recommended.

But Megan’s history of depression falls into the “severe” category, and a prior episode of depression did not respond well to psychotherapy. Your recommendation to her, therefore, is that she should continue taking an antidepressant—unless she feels strongly that she should discontinue it.

Risks of SRIs in pregnancy

Megan considers what you’ve discussed about her high risk of developing recurrent depression during pregnancy. She decides that she wants to continue taking her antidepressant during pregnancy, but she has concerns—based on what she has been reading on the Internet.

Megan hands you a detailed printout downloaded from a Web site unfamiliar to you and asks about risks to the baby of such medications as paroxetine.

What should you tell Megan about SRIs in pregnancy—paroxetine, specifically?

You preface your remarks to her by noting that the data physicians work with are imperfect—because randomized, controlled clinical trials pose an ethical dilemma as a method of study in pregnant women. You then discuss with her current scientific understanding of potential risks to her fetus.

The difference in the rates of structural malformation among SRI-exposed and SRI-unexposed groups has been studied; most studies have found no increased rate of major or specific cardiac malformations.¹² However, first-trimester paroxetine appeared, in some studies, to be associated with an increased rate of cardiac malformations. That led to a category-“D” pregnancy classification in 2005 and an FDA “Public Health Advisory.”

Other large cohort studies have not uncovered such an association. It has been hypothesized that the methodology of data collection may have influenced this finding.¹³

Other malformations have been implicated in some studies but not others, and have included associations between specific SRIs and cardiac ventricular outflow defects, craniosynostosis, and omphalocele. The absolute risk of these defects remains extremely low, however, and close to the background rate seen in the general population.¹⁴

Megan asks: “With that risk-category ‘D’ for paroxetine, do you recommend I continue taking it or should I switch to another medication while I’m pregnant?”

You review again with Megan that, although some studies have linked first-trimester paroxetine to an increased risk of cardiac malformation, that finding has not been replicated in several large cohort studies. You explain that, if she had a history of recurrent depression that had failed to respond to many antidepressants and only paroxetine worked, an attempt at switching the SRI would not be recommended because of the potential for relapse.

Megan tells you that she would feel safer not taking a category-“D” drug. You agree and propose a judicious approach: Because she has come to see you before she became pregnant, with enough time to complete a slow crossover to an alternative SRI, and because she has not had any earlier trials of other SRIs, a slow taper of paroxetine, coupled with a crossover to an alternative SRI, is a reasonable option—with the caution that substitution always carries a risk of relapse.

Problems in newborns

Megan considers the risks you’ve discussed so far. She remembers a recent article in a magazine for pregnant women that described severe “respiratory” and “withdrawal” symptoms in infants who were born to mothers taking an SRI antidepressant. She wonders if she should consider discontinuing her SRI in the third trimester to try to mitigate those risks.

Megan is asking you about an SRI exposure risk that has been fairly consistent across studies, called neonatal abstinence syndrome (NAS) or poor neonatal adaptation.

NAS is a cluster of symptoms that occurs in 15% to 30% of newborns who have been exposed to an SRI during the third trimester of pregnancy.¹⁵ Signs include irritability, weak cry, tachypnea, temperature instability, and hypoglycemia—all of which are transient, peak during the first 48 hours after delivery, and resolve in less than 2 weeks.

Multiple hypotheses have been put forward to account for NAS, including the possibilities that it reflects a withdrawal syndrome, pharmacotoxicity, or an underlying gene–SRI interaction. The physiology behind NAS remains unknown, however.¹²

Megan next asks you about persistent pulmonary hypertension of the newborn (PPHN). You explain that PPHN is of recent concern in women who have been taking an SRI in the latter half of their pregnancy.

The rate of PPHN in the general population is 0.5 to 2 newborns for every 1,000. Associated mortality is approximately 10% to 20%.^16-18 This rate is thought to rise to approximately 6 of every 1,000 newborns among those who have been exposed to an SRI in utero—with some evidence of increased risk conferred through SRI exposure during later pregnancy (studies define this as the second half of the pregnancy).¹⁵ Although the relative risk of PPHN is increased threefold to sixfold when an SRI is used in pregnancy, absolute risk remains extremely low.

Concerns have been raised over research methodology in the few studies that have looked into SRI exposure and PPHN. Not all such studies found a change in relative risk or absolute risk of PPHN in SRI- exposed infants, compared to what was found in non-SRI–exposed infants.^15,19,20

Megan presses you, however, with the understandable question of whether she should taper her SRI during the last trimester (which the Web site she has found recommends). With the above information in mind, you explain that, given current understanding of the low absolute risk of PPHN, and given her illness history and severity of prior depression, you would not recommend that she taper the antidepressant in the third trimester.

Furthermore, the same counsel applies in regard to NAS: Given the risk of psychiatric morbidity caused by discontinuing an SRI during the third trimester, you do not recommend that she taper an SRI during that period to avoid NAS.

You explain that, instead, physicians now counsel women who take an SRI about the signs of NAS so that they can be prepared if they observe any of them in their infant.

Megan has one more question: “Will I be able to breastfeed while I’m taking an antidepressant?”

Postpartum issues to consider

Given the inherent difficulties and risks of relapse associated with a crossover to an alternative antidepressant postpartum, it makes sense, when possible, for a woman to take an antidepressant during pregnancy that can safely be continued while she is breastfeeding.

You tell Megan that, even though the quality of the data in this area is also thin, SRIs that have a low maternal serum profile are considered safest in breastfeeding.

To date, two SRIs—sertraline and paroxetine—have not been detectable in the breast milk of women taking either of them.²¹

CASE Appointment concluded, overflowing with information,
advice, and optimism

Megan says that, taking into account all that you and she have talked about, and even though she wants to return with her husband, she would like to switch to sertraline before she becomes pregnant—while she gauges its effectiveness at keeping her disorder in remission.

A good outcome requires you to prevail over obstacles

Because a diagnosis of depression spans a continuum of severity and, often, is not perceived as an acutely life-threatening illness, evaluating the risks and benefits of treatment is a murky undertaking.

Our role as physicians is to, first, educate ourselves and our patients about these variables and, second, support our patients in the decisions that they make. Physicians who care for pregnant women must be aware of the benefits and limitations of treatments as reported in the most current literature if they are going to assist women with decisions about treatment in the best possible way.

Social stigma. There remains the impact of stigma. Depressive and anxiety disorders are often perceived to be either under the control of an affected person’s “free will” or not as serious as other forms of “medical” disease. Consequently, the role that cultural and social pressures play in the risk–benefit analysis conducted by pregnant women and their physicians can’t be discounted.

Customized decision-making. As more data emerge about the treatment of depression in pregnancy, it has become clear: Treatment algorithms meant to simplify our decisions must always be individualized and extended into the postpartum period.

Treatment selection. Management of mild depression during pregnancy does not always require medication. Multiple variables—the list is long, and includes a patient’s psychiatric history, family psychiatric history, response to prior treatment, severity of depression, severity of prior depression, degree of social support, and personal desires—must be considered in determining what treatment is appropriate before, during, and after a pregnancy.

For a woman who suffers mild or moderate depression, with few antenatal depression risk factors, a trial of psychotherapy is recommended as first-line treatment. For a woman suffering from severe depression, or one who has a history of severe depression that has not responded well to psychotherapy alone, continuation or initiation of an SRI antidepressant is the current recommendation.

We want to hear from you! Tell us what you think.

CAS: Depressive disorder, anticipating a pregnancy

Your patient Megan—well-educated, 29 years old, G0P0—has come to you to discuss her antidepressant (paroxetine [Paxil]) because she is planning her first pregnancy.

Megan has a history of recurrent major depressive disorder (MDD), which is in remission (see “What is MDD?”).

How will you begin the conversation with this patient about keeping MDD in remission during her pregnancy and ensuring the safety of her fetus?

There is a 20% to 25% lifetime prevalence of depression in women; the disorder peaks during childbearing years, however.¹ As of 2003, 13% of pregnant women had taken an antidepressant at some time during their pregnancy, a percentage that has doubled since it was assessed in 1999.²

You are faced with several quandaries in deciding whether to recommend that your patient continue, or discontinue, antidepressant therapy during pregnancy:

• As many as 68% of women who terminate antidepressant treatment before or during pregnancy relapse.
• Even 26% of women who continue antidepressant during pregnancy relapse—requiring a dosage adjustment or change in treatment.³
• Yet the possibly elevated cortisol levels of severe, untreated depression may harm the placenta and fetus.^4,5

So, what do you need to know to assess the risks and benefits of “Megan” stopping, or continuing, paroxetine during her anticipated pregnancy? And what are the risks to Megan’s fetus of treating, or not treating, her depression with a serotonin reuptake inhibitor (SRI*)?

Gauging the risks of depression in pregnancy

In any given patient, her history and family history of depression are key to determining the likelihood that she will suffer ongoing or recurrent depression.

CASE continued Repeated treated episodes plus a family history

In obtaining Megan’s history, you learn that she has had three prior episodes of depression, all of which were successfully treated with paroxetine. Megan has been stable on paroxetine for 3 years.

Notably, the second episode of depression was initially treated with a 16-week trial of psychotherapy alone; when depressive symptoms did not remit, paroxetine was added. That episode was considered severe because it included pervasive thoughts of suicide.

You also learn that Megan’s mother suffered from postpartum depression and that her father and paternal grandmother were treated for depression.

Known risk factors for depression during pregnancy include: maternal anxiety; prior diagnosis of depression during pregnancy; history of postpartum anxiety or depression; prior diagnosis of either anxiety or depressive disorder; significant life stress (e.g., divorce, death of a loved one); degree of social support—particularly, intimate social support; “intendedness” of pregnancy; domestic violence; and insurance status.⁶

You review with Megan her risk factors for depression during pregnancy, namely: three prior episodes of MDD and a strong family history of mood disorder. Her MDD is considered “severe” because she has a history of suicidality. You tell Megan that, given these factors, she is at high risk of a recurrence of her depressive illness during pregnancy.

Megan asks: “Would getting depressed during pregnancy hurt the baby?”

Depression during pregnancy affects both infant and maternal well-being, although studies are in conflict about the extent of that morbidity. Multiple areas of potential risk to mother and infant have been studied, including the effect of depression on:

• maternal well-being
• growth of the infant
• spontaneous abortion
• preterm delivery
• neonatal physiologic and neurobehavioral measures
• long-term considerations for the developing infant and child.

Within these categories of risk, a diagnosis of depression during pregnancy has been associated (in some but not all studies) with a higher risk, or rate, of:

• postpartum depression
• preterm birth
• lower maternal weight gain
• maternal tobacco, alcohol, and other substance use
• lower infant gestational age at birth
• small-for-gestational age infant birth.^7-10

In terms of long-term impact on offspring, studies differ in their estimation of risk; however, children exposed to untreated, maternal depression at 18 weeks’ and 32 weeks’ gestation did show a greater degree of developmental delay at 18 months than children who were born to a mother who was not depressed during pregnancy.¹¹

You discuss these risks with Megan. She asks: “What treatment do you recommend for me?” You turn to the 2009 guidelines published jointly by the American Psychiatric Association (APA) and ACOG.

These guidelines recommend that you consider 1) the severity of her current depression, 2) her history of depression severity, and 3) her preference for treatment.¹² For mild depression during pregnancy, when there is no history of severe depression, or for a history of depression that responded well to psychotherapy in the past, a trial of psychotherapy without medications is recommended.

But Megan’s history of depression falls into the “severe” category, and a prior episode of depression did not respond well to psychotherapy. Your recommendation to her, therefore, is that she should continue taking an antidepressant—unless she feels strongly that she should discontinue it.

Risks of SRIs in pregnancy

Megan considers what you’ve discussed about her high risk of developing recurrent depression during pregnancy. She decides that she wants to continue taking her antidepressant during pregnancy, but she has concerns—based on what she has been reading on the Internet.

Megan hands you a detailed printout downloaded from a Web site unfamiliar to you and asks about risks to the baby of such medications as paroxetine.

What should you tell Megan about SRIs in pregnancy—paroxetine, specifically?

You preface your remarks to her by noting that the data physicians work with are imperfect—because randomized, controlled clinical trials pose an ethical dilemma as a method of study in pregnant women. You then discuss with her current scientific understanding of potential risks to her fetus.

The difference in the rates of structural malformation among SRI-exposed and SRI-unexposed groups has been studied; most studies have found no increased rate of major or specific cardiac malformations.¹² However, first-trimester paroxetine appeared, in some studies, to be associated with an increased rate of cardiac malformations. That led to a category-“D” pregnancy classification in 2005 and an FDA “Public Health Advisory.”

Other large cohort studies have not uncovered such an association. It has been hypothesized that the methodology of data collection may have influenced this finding.¹³

Other malformations have been implicated in some studies but not others, and have included associations between specific SRIs and cardiac ventricular outflow defects, craniosynostosis, and omphalocele. The absolute risk of these defects remains extremely low, however, and close to the background rate seen in the general population.¹⁴

Megan asks: “With that risk-category ‘D’ for paroxetine, do you recommend I continue taking it or should I switch to another medication while I’m pregnant?”

You review again with Megan that, although some studies have linked first-trimester paroxetine to an increased risk of cardiac malformation, that finding has not been replicated in several large cohort studies. You explain that, if she had a history of recurrent depression that had failed to respond to many antidepressants and only paroxetine worked, an attempt at switching the SRI would not be recommended because of the potential for relapse.

Megan tells you that she would feel safer not taking a category-“D” drug. You agree and propose a judicious approach: Because she has come to see you before she became pregnant, with enough time to complete a slow crossover to an alternative SRI, and because she has not had any earlier trials of other SRIs, a slow taper of paroxetine, coupled with a crossover to an alternative SRI, is a reasonable option—with the caution that substitution always carries a risk of relapse.

Problems in newborns

Megan considers the risks you’ve discussed so far. She remembers a recent article in a magazine for pregnant women that described severe “respiratory” and “withdrawal” symptoms in infants who were born to mothers taking an SRI antidepressant. She wonders if she should consider discontinuing her SRI in the third trimester to try to mitigate those risks.

Megan is asking you about an SRI exposure risk that has been fairly consistent across studies, called neonatal abstinence syndrome (NAS) or poor neonatal adaptation.

NAS is a cluster of symptoms that occurs in 15% to 30% of newborns who have been exposed to an SRI during the third trimester of pregnancy.¹⁵ Signs include irritability, weak cry, tachypnea, temperature instability, and hypoglycemia—all of which are transient, peak during the first 48 hours after delivery, and resolve in less than 2 weeks.

Multiple hypotheses have been put forward to account for NAS, including the possibilities that it reflects a withdrawal syndrome, pharmacotoxicity, or an underlying gene–SRI interaction. The physiology behind NAS remains unknown, however.¹²

Megan next asks you about persistent pulmonary hypertension of the newborn (PPHN). You explain that PPHN is of recent concern in women who have been taking an SRI in the latter half of their pregnancy.

The rate of PPHN in the general population is 0.5 to 2 newborns for every 1,000. Associated mortality is approximately 10% to 20%.^16-18 This rate is thought to rise to approximately 6 of every 1,000 newborns among those who have been exposed to an SRI in utero—with some evidence of increased risk conferred through SRI exposure during later pregnancy (studies define this as the second half of the pregnancy).¹⁵ Although the relative risk of PPHN is increased threefold to sixfold when an SRI is used in pregnancy, absolute risk remains extremely low.

Concerns have been raised over research methodology in the few studies that have looked into SRI exposure and PPHN. Not all such studies found a change in relative risk or absolute risk of PPHN in SRI- exposed infants, compared to what was found in non-SRI–exposed infants.^15,19,20

Megan presses you, however, with the understandable question of whether she should taper her SRI during the last trimester (which the Web site she has found recommends). With the above information in mind, you explain that, given current understanding of the low absolute risk of PPHN, and given her illness history and severity of prior depression, you would not recommend that she taper the antidepressant in the third trimester.

Furthermore, the same counsel applies in regard to NAS: Given the risk of psychiatric morbidity caused by discontinuing an SRI during the third trimester, you do not recommend that she taper an SRI during that period to avoid NAS.

You explain that, instead, physicians now counsel women who take an SRI about the signs of NAS so that they can be prepared if they observe any of them in their infant.

Megan has one more question: “Will I be able to breastfeed while I’m taking an antidepressant?”

Postpartum issues to consider

Given the inherent difficulties and risks of relapse associated with a crossover to an alternative antidepressant postpartum, it makes sense, when possible, for a woman to take an antidepressant during pregnancy that can safely be continued while she is breastfeeding.

You tell Megan that, even though the quality of the data in this area is also thin, SRIs that have a low maternal serum profile are considered safest in breastfeeding.

To date, two SRIs—sertraline and paroxetine—have not been detectable in the breast milk of women taking either of them.²¹

CASE Appointment concluded, overflowing with information,
advice, and optimism

Megan says that, taking into account all that you and she have talked about, and even though she wants to return with her husband, she would like to switch to sertraline before she becomes pregnant—while she gauges its effectiveness at keeping her disorder in remission.

A good outcome requires you to prevail over obstacles

Because a diagnosis of depression spans a continuum of severity and, often, is not perceived as an acutely life-threatening illness, evaluating the risks and benefits of treatment is a murky undertaking.

Our role as physicians is to, first, educate ourselves and our patients about these variables and, second, support our patients in the decisions that they make. Physicians who care for pregnant women must be aware of the benefits and limitations of treatments as reported in the most current literature if they are going to assist women with decisions about treatment in the best possible way.

Social stigma. There remains the impact of stigma. Depressive and anxiety disorders are often perceived to be either under the control of an affected person’s “free will” or not as serious as other forms of “medical” disease. Consequently, the role that cultural and social pressures play in the risk–benefit analysis conducted by pregnant women and their physicians can’t be discounted.

Customized decision-making. As more data emerge about the treatment of depression in pregnancy, it has become clear: Treatment algorithms meant to simplify our decisions must always be individualized and extended into the postpartum period.

Treatment selection. Management of mild depression during pregnancy does not always require medication. Multiple variables—the list is long, and includes a patient’s psychiatric history, family psychiatric history, response to prior treatment, severity of depression, severity of prior depression, degree of social support, and personal desires—must be considered in determining what treatment is appropriate before, during, and after a pregnancy.

For a woman who suffers mild or moderate depression, with few antenatal depression risk factors, a trial of psychotherapy is recommended as first-line treatment. For a woman suffering from severe depression, or one who has a history of severe depression that has not responded well to psychotherapy alone, continuation or initiation of an SRI antidepressant is the current recommendation.

We want to hear from you! Tell us what you think.

Celiac disease (CD), also known as gluten-sensitive enteropathy or celiac sprue, is an endocrine disorder whose effects are triggered by the ingestion of gluten—the principle storage protein in wheat, rye, and barley.^1-3 CD inflicts damage to the mucosa of the small intestine and subsequently to systemic organ tissues. CD can affect any organ in the body.¹ The responsible genetic factors are the human leukocyte antigens, HLA -DQ2 and -DQ8, which are present in 40% of the general population but are found in nearly 100% of patients with CD.^1,3,4

Though previously considered uncommon, CD has been estimated to affect more than 1% of the general population worldwide.^1,4,5 Currently, CD is most reliably identified by ­positive serum antibodies, specifically immunoglobulin A (IgA) anti-tissue transglutaminase (tTG) and IgA antiendomysial (EMA) antibodies,⁶ and by a finding of villous atrophy of the intestinal lining on biopsy. The spectrum of presentations of CD is broad, including the “typical” intestinal features of diarrhea, bloating, abdominal pain, and weight loss; or common “atypical” extraintestinal manifestations, such as anemia, osteoporosis, infertility, and neurologic disturbances (eg, peripheral neuropathy⁷).⁸ See Table 1.^8,9

Prevalence of CD is greater among those with a family history of CD; with autoimmune diseases, especially type 1 diabetes mellitus (T1DM) and thyroiditis; and with certain genetic disorders (ie, Down, Turner, and Williams syndromes).^8-15 Because atypical features dominate in older children and adults, many cases escape diagnosis, and patients may be exposed to serious long-term complications, such as infertility and cancer.¹

CD is a lifelong condition, necessitating the complete exclusion of gluten-containing products from the diet. In the US food industry, gluten is used in numerous food applications, complicating the patient education and lifestyle changes needed to implement and maintain a gluten-free diet (GFD). However, if a GFD is not strictly followed, the patient’s quality of life can be seriously impaired.^1,4,5

AWARENESS ESSENTIAL IN PRIMARY CARE

For the primary care provider (PCP), there is no shortage of patients with gastrointestinal (GI) disorders, thyroid disease, diabetes, anemia, fatigue, or dysmenorrhea; additionally, PCPs regularly treat patients for a number of associated disorders, including anxiety, irritability, and attention deficit. Yet how likely are PCPs to screen patients with these symptoms for CD? And how many patients with CD never receive a diagnosis of the disorder?

In fact, it has been estimated that more than 90% of persons affected by CD are currently undiagnosed.^1,4 In one study involving mass screening of 1,000 children ages 2 to 18, it was determined that almost 90% of celiac-positive children had not previously been diagnosed.⁵ Similarly, in a cohort of 976 adults (median age, 54.3), the diagnostic rate for CD was initially low at 0.27 cases per thousand visits but increased to 11.6 cases per thousand visits after implementation of active screening.⁴ Based on these data, it has been estimated that more than 2.7 million Americans unknowingly carry this potentially life-threatening genetic disease.^1,4,16  

Given the potential patient population with undetected, untreated CD, some researchers consider the disorder one of the most common lifelong diseases in the US.^1,8,16 CD is closely associated with T1DM and autoimmune thyroiditis, with cross- prevalence at 11% and 6.7%, respectively.^8,12,17 The close association between T1DM and CD led the American Diabetes Association¹⁸ to amend guidelines in 2009, suggesting screening for CD in all patients newly diagnosed with T1DM.

PATIENT PRESENTATION: ADULTS VERSUS CHILDREN

Most infants and young children with CD present with the typical or “classic” triad of signs: short stature, failure to thrive, and diarrhea; in individual patients, however, the impact of genetics and exposure to gluten over time can cause considerable variation in patient presentation. As patients with undiagnosed CD age, they may present quite differently or even revert to a latent stage and become asymp­tomatic.^19,20

In two separate reviews, it was noted that classic symptoms of CD are not evident in a majority of older children and adults; instead, anemia and fatigue were the predominating symptoms.^12,20An important note: The patient with no symptoms or atypical signs of CD may still be experiencing significant damage, inflicted by gluten-induced antibodies, to the intestinal lining and/or mucosal linings in other organ systems—perhaps for years before the disease becomes evident.²⁰ 

Clinical Findings Differ With Age, Gender

Historically, CD was considered a pediatric syndrome; however, a diagnosis of CD has become increasingly common among older children and adults, especially elderly patients, although symptoms in the latter group are subtle.^21-23 Recent, active CD is being diagnosed among men older than 55 more commonly than in women of this age-group; women are generally younger at diagnosis but have experienced symptoms longer.^22,23 This later onset in men suggests that antibody seropositivity and the associated active disease may be triggered later in life.^8,22

A variety of findings have been reported in the history and physical exam of most patients who present with CD.The most prevalent signs and symptoms are abdominal pain, frequent loose stools, weight loss, joint pain, and weakness.^8,11,16 Unlike the pediatric patient with the classic triad of symptoms, adults usually experience more generalized GI manifestations, such as irritable bowel syndrome (IBS), abdominal pain, or acid reflux.¹⁰

Many patients have no GI symptoms but may present solely with fatigue, arthralgias, or myalgia.²⁰ In fact, more than 50% of adults with CD present with atypical or extraintestinal disorders, such as anemia, infertility, osteoporosis, neurologic problems, or other autoimmune disorders.^8,16,23,24 It is important for clinicians to note that atypical is somewhat typical in the older patient who presents with CD.

Patients with asymptomatic or silent CD, (see “Classification and Pathology,” below) lack both classic and atypical symptoms but still have villous atrophy, usually discovered during endoscopy being conducted for other reasons.⁸ Because of its predominantly atypical presentations, CD is considered a multisystem endocrine condition rather than one that is mainly gastrointestinal.^8,16,25,26

CLASSIFICATION AND PATHOLOGY

Though frequently a silent disorder, CD typically progresses through four stages: classical, atypical, latent, and silent. Clinicians should strive to become fully aware of each stage and its implications.^8,26,27

The classical form is primarily diagnosed in children ages 6 to 18 months. It is characterized by villous atrophy and typical symptoms of intestinal malabsorption.⁸

The patient with atypical CD has minor intestinal symptoms, but architectural abnormalities can be found in the mucosa of the small intestine. This patient is likely to present with various extraintestinal disorders, including osteoporosis, anemia, infertility, and neuropathies.^7,8

In the latent form of CD, the HLA-DQ2 and/or -DQ8 genetic markers are present. Serology for CD may be positive, but the intestinal mucosa may be normal. The patient may or may not be experiencing extraintestinal symptoms. In patients with latent CD, the gluten-associated changes appear later in life.⁸ The precise trigger for late activation of the disease, though apparently linked to genetics and gluten exposure, remains elusive.^20,24

The silent form of CD is marked by mucosal abnormalities in the small intestine and usually by positive CD serology, but it is asymptomatic. The iceberg theory of celiac disease²⁸ (see figure²⁸) has been proposed to explain CD’s hidden manifestations over time.

In patients with atypical, latent, or silent CD, the condition is sometimes detected incidentally during screening of at-risk groups or by endoscopy performed for other reasons.⁸ Most of these patients respond well to GFD therapy, noting both physical and psychological improvement—suggesting that these patients, even though asymptomatic and seemingly healthy, may have been experiencing minor manifestations of undiagnosed CD for many years: decreased appetite, fatigue, and even behavioral abnormalities.^1,8

Histopathologic analysis of abnormalities found on biopsy of the small intestine relies on the four-stage Marsh classification²⁹:

• Marsh 0: normal mucosa

• Marsh I: intraepithelial lymphocytosis

• Marsh II: intraepithelial lymphocytosis with crypt hyperplasia

• Marsh III: intraepithelial lymphocytosis with crypt hyperplasia and villous atrophy.^8,29 Modifications to this classification have been made by Oberhuber^30,31 to denote the degree of villous flattening³² (ie, IIIa, IIIb, IIIc).

Villous atrophy of the mucosa has long been considered the hallmark of CD, and its detection, according to the American Gastroenterological Association,^2,26 remains the gold standard in confirming a diagnosis of CD.^4,16,26 However, early screening (ie, serologic testing for tTG and EMA) is the necessary initial step in ensuring diagnostic accuracy, as other conditions can cause villous atrophy, and latent CD can coexist with normal intestinal mucosa.¹⁰

Avoiding Diagnostic Delays

Because of the broad spectrum of unrelated GI signs in all ages and the subtle presentation in adults, diagnosis of CD in this patient population is frequently delayed for estimated periods ranging from five to 11 years.^4,11,23,33

Improving clinician awareness of the manifestations of CD is essential³⁴; too frequently, the common symptoms of probable CD are treated as individual idiopathic disorders by both PCPs and secondary specialists, who prescribe proton pump inhibitors, antihistamines, cathartics, and/or antimotility drugs for years without ruling out a common, easily identified genetic disease. Even though the prevalence of CD has recently been shown to have increased more than fourfold since 1950,³⁵ serologic testing for CD is not widely implemented by PCPs.^4,11,20

Specialists, too, may be slow to recognize this treatable autoimmune disorder. In a recent nationwide study, it was found that gastroenterologists performed a small-bowel biopsy in less than 10% of their patients who underwent esophagogastroduodenoscopy (EGD) for likely symptoms of CD.³⁶ Relying solely on clinical expertise and visual recognition of intestinal abnormalities can delay diagnosis for years.^4,36 Many patients may never be given a correct diagnosis of CD.

The Role of Serologic Testing

Current data demonstrate that autoimmune diseases are on the rise,^8,16,36 and CD can be the primary cause or a contributing factor in several other disorders (see Table 2^8,16,37,38). Gastroenterologists may be correct in stating that biopsy is the only way to make a diagnosis of CD or to stage CD-associated intestinal damage^4,26; yet by implementing a protocol of serologic testing for tTG and EMA in at-risk patients, PCPs could prevent a missed diagnosis on EGD when biopsy has not been considered, as in the case of atypical CD; or when biopsy results are negative in a patient with latent CD.^39,40

Because of its high negative predictive value, serologic testing should be conducted first to significantly reduce the probability of suspected CD. Such selective screening should be performed by the PCP before invasive testing by the gastroenterologist and before long-term empiric treatment for idiopathic GERD, IBS, or other unexplained disorders.^32,40

Thus, it has been recommended that PCPs perform screening for CD in patients with unexplained chronic GI disturbances or a familial prevalence of CD, or in those who present with the atypical signs of CD or with associated disorders.^1,10,16,20 Whether serologic screening results are positive or negative for CD, the patient with classic GI symptoms should undergo endoscopy with biopsy to confirm active disease and to evaluate the extent of intestinal damage—or to explore other causes.^26,39 An algorithm^{2,4,8,11,16,26} illustrating suggested screening, treatment, and follow-up strategies for patients at high risk for CD is shown below.

Catassi and Fasano³⁴ recently proposed a “four out of five” rule, by which diagnosis of CD may be confirmed in patients with at least four of the following five criteria:

• Typical symptoms of CD

• Positive serology (ie, IgA tTG and IgA EMA antibodies)

• Genetic susceptibility (as confirmed by the presence of HLA-DQ2 and HLA-DQ8)

• Small intestine biopsy results indicating celiac enteropathy

• Improvement of CD signs and symptoms following implementation of the GFD.³⁴

CURRENT TREATMENT AND ASSOCIATED CHALLENGES

Because gluten consumption is the principal trigger of CD pathology, a GFD is considered the safest, most effective therapy for the disorder.^{1,8,10,11,16,19} Implementing and maintaining the GFD involves a considerable learning curve for the patient, the patient’s family, and possibly the provider; to achieve complete recovery, all involved must become knowledgeable regarding gluten-free and gluten-containing products. The patient must be willing and able to avoid those that contain gluten and bear the potentially high costs⁸ of gluten-free foods.

Even for patients with CD who are determined to comply with the GFD, gluten monitoring can be difficult. There are ways to determine what is a safe level of gluten ingestion for each patient, but trace amounts of gluten are found in many products, including some that are marked “gluten-free.”^1,41 The FDA has proposed that a product labeled gluten-free may contain no more than 20 parts per million (ppm, ie, 20 mg/kg) of gluten.⁴² In other countries, however, acceptable levels may be as high as 200 ppm (200 mg/kg)—which are considered well above the trigger amounts in the average patient with CD.^1,41 The complex nature of each patient’s sensitivity to gluten and the ubiquitous presence of gluten as a food source in both industrialized and developing countries make adherence to the GFD challenging.¹⁰

It is critical for the PCP to help the patient review all of his or her prescription and OTC pharmaceuticals and nutritional supplements, as these may contain hidden gluten in the form of modified starches and other fillers.⁴¹ It may be also advisable to involve the patient’s pharmacist, requesting an assessment for agents that may be suspect.

A management team approach may ensure the most integrative care. In addition to the PCP and the pharmacist, such a team might include a gastroenterologist, an endocrinologist, a nutritionist, and a psychologist, who may be needed to help the patient confront the great life adjustment required, in addition to addressing other behavioral disorders that are common in patients with CD.^10,26

See the box for resources that may be beneficial for both patients and their clinicians.

Alternative Medicine Options

Alternative medicine is gaining favor, especially when no drug therapy is currently available to alleviate gluten toxicity. Supplementation with the fat-soluble vitamins (A, D, E, and K), vitamin B₁₂, folic acid, and the minerals calcium and iron, as indicated by serum deficiencies, is recommended.^10,20 Supplementation with digestive enzymes, which are known to be deficient in patients with CD as a result of villous atrophy, may help break down undigested gluten proteins; research is under way to find a recombinant enzyme therapy.¹⁰ Researchers have recently shown that probiotics (specifically, Bifidobacterium lactis) significantly reduce the immune response when incidental exposure to gluten occurs.⁴³ 

REFRACTORY CELIAC DISEASE

Patients with late-onset CD, especially those not diagnosed until after age 50, may have a diminished or absent response to dietary therapy. In some patients, histologic signs and clinical symptoms persist or relapse after a prior positive response to a strict GFD, despite continued adherence to the diet for longer than 12 months.⁴⁴ Once other causes have been carefully excluded, these patients are considered to have refractory celiac disease (RCD). Exact prevalence of RCD is unknown, but Tack et al⁸ estimate it at 5% of all cases of CD. Relapsing CD resulting from poor adherence to the GFD is not considered true RCD.^8,16,37,38

According to researchers for the European Celiac Disease working group,^8,45 RCD can be divided into types I and II:

• RCD I, in which normal polyclonal T cells are present in the intestinal lumen

• RCD II, in which abnormal clonal T cells infiltrate the intestinal mucosa, representing premalignancy.⁴⁵

The histologic picture of RCD mimics that of severe CD. Malabsorption complications, lesions in the intestinal mucosa, and inflammatory lymphocytosis are present.⁴⁴ Some patients, like those with classical CD, have serology test results that are consistent with CD and an initial response to GFD therapy; after months or years, however, this response subsides. Other patients are immediately unresponsive to GFD and lack the serologic markers for CD.⁸

A differential diagnosis including other explanations for the manifestations of RCD must be carefully reviewed, with each excluded, through the strategies shown in Table 2. This review is essential, as patients with RCD II have a much worse prognosis than those with RCD I; the associated five-year survival rates are 44% to 58%, versus 85% or greater, respectively.^36,46

Additionally, the continued autoimmune expansion of aberrant T cells in patients with RCD II causes early conversion to malignancy, usually within four to six years after diagnosis. Enteropathy-associated T-cell lymphoma is the most common malignancy, occurring in more than 50% of patients with RCD II, and a likely cause of death.^3,8,46,47

Treatment for Refractory Celiac Disease

In addition to the GFD, patients with RCD I generally respond well to corticosteroids or other immunosuppressive treatment.⁸ Use of budesonide, a corticosteroid given in a once-daily, 9-mg dose, has led to almost complete recovery in most patients. Duration of therapy is response-dependent.³⁷

Systemic corticosteroids or other immunosuppressant agents, such as azathioprine, should be reserved for patients with RCD I or RCD II who do not respond to budesonide, as lengthy treatment regimens are required, with considerable risk for adverse effects.^35,48

Recently, promising results have been reported in a small, open-label cohort study involving patients with RCD II who underwent five days of treatment with IV cladribine (0.1 mg/kg/d).⁴⁹

PREVENTION OF CD

A good nutritional start from birth could be the best means of preventing symptomatic CD. According to findings from a meta-analysis of data from four studies, children being breastfed at the time gluten was introduced had a 52% reduction in risk for CD, compared with their peers who were not being breastfed at that time.⁵⁰

The protection breast milk appears to provide against CD is not clearly understood. One possible mechanism is that breast milk may protect an infant against CD by preventing gastrointestinal infections, as is the case with other infections. The presence of GI infections (eg, rotavirus) in early life could lead to increased permeability of the intestinal mucosa, allowing the passage of gluten into the lamina propria.^3,8,50

Extended duration of breastfeeding is also associated with a reduced risk for CD.^8,41,50 Long-term studies are needed, however, to determine whether breastfeeding delays CD onset or provides permanent protection against the disorder.

RECENT DEVELOPMENTS

A recently marketed OTC testing kit for CD is now available in Canada and other countries outside the US; this may be an indication of the growing awareness of the numbers of patients with undiagnosed CD. The test parallels the tTG serum test, which in the US is evaluated only in laboratories; it has comparable specificity and sensitivity, with results within 10 minutes. In the US, the FDA has not yet approved the kit, but domestic testing of the product may soon be under way.⁵¹

Alternative treatment modalities are currently focusing on the detoxification of wheat components, rapid enzymatic degradation to reduce exposure to gluten, inducing gluten tolerance, inhibiting permeability of the small intestine to gluten (which, it is thought, may prevent many of the systemic manifestations of CD), and finally, development of an immunomodulatory vaccine.^8,33 None of these therapies is yet approved.

IMPLICATIONS OF DELAYED DIAGNOSIS

The unrecognized prevalence of CD is a growing issue, as many symptomatic but unscreened patients are frequently misdiagnosed with IBS, chronic fatigue, or other idiopathic disorders. The silent and latent forms of CD are of the greatest concern, as they show minimal signs and can lead to multiple organ system damage and are implicated in other autoimmune disorders. The longer diagnosis is delayed, the greater is patients’ resistance to dietary therapy, and the less likely that established intestinal and/or neurologic damage can be reversed.^10,20,51

The large proportion of undiagnosed celiac patients may account for an accompanying underestimated cost to both the patient and the health care system because of repeated referrals to investigate unexplained disorders before an accurate diagnosis is made. In one recent analysis, mass screening for CD in a young adult population led to improved quality-of-life years by shortening the time to diagnosis and treatment; it was also found cost-effective.⁵² PCPs must be attentive to patients who may be at high risk for CD and implement combined serum tTG and EMA screening as the initial step in identification and treatment.^4,10,11,20 Some form of standardized screening protocol may become inevitable.

CONCLUSION

The prevalence of CD has increased more than fourfold since 1950, and diagnosis is often significantly delayed. Increased awareness is needed among PCPs that CD in adults is likely to manifest with atypical (ie, nongastrointestinal) symptoms and signs. Judicious use of serologic screening for CD would lead to earlier diagnosis and more effective treatment, possibly preventing the potentially lethal refractory disease forms associated with chronic untreated CD.  

REFERENCES

1. Ramos M, Orozovich P, Moser K, et al. Health 1. Catassi C, Fasano A. Celiac disease. Curr Opin Gastroenterol. 2008;24(6):687-691.

2. AGA Institute. AGA Institute medical position statement on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1977-1980.

3. Green PH, Jabri B. Coeliac disease. Lancet. 2003;362(9381):383-391.

4. Catassi C, Kryszak D, Louis-Jacques O, et al. Detection of celiac disease in primary care: a multicenter case-finding study in North America. Am J Gastroenterol. 2007;102(7);1454-1460.

5. Demirçeken FG, Kansu A, Kuloglu Z, et al. Human tissue transglutaminase antibody screening by immunochromatographic line immunoassay for early diagnosis of celiac disease in Turkish children. Turk J Gastroenterol. 2008;19(1):14-21.

6. van der Windt DA, Jellema P, Mulder CJ, et al. Diagnostic testing for celiac disease among patients with abdominal symptoms: a systematic review. JAMA. 2010;303(17):1738-1746.

7. Freeman HJ. Neurological disorders in adult celiac disease. Can J Gastroenterol. 2008; 22(11):909-911.

8. Tack GJ, Verbeek WHM, Schreurs MWJ, Mulder CJJ. The spectrum of celiac disease: epidemiology, clinical aspects and treatment. Nat Rev Gastroenterol Hepatol. 2010;7(4):204-213.

9. Farrell RJ, Kelly CP. Celiac sprue. N Engl J Med. 2002;346(3):180-188.

10. Green PHR, Cellier C. Celiac disease. N Engl J Med. 2007;357(17):1731-1743.

11. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003;163(3):286-292.

12. Sud S, Marcon M, Assor E, et al. Celiac disease and pediatric type 1 diabetes: diagnostic and treatment dilemmas. Int J Pediatr Endocrinol. 2010;2010:161285. Epub 2010 Jun 23.

13. Swigonski NL, Kuhlenschmidt HL, Bull MJ, et al. Screening for celiac disease in asymptomatic children with Down syndrome: cost-effectiveness of preventing lymphoma. Pediatrics. 2006;118(2):594-602.

14. Bonamico M, Pasquino AM, Mariani P, et al; Italian Society of Pediatric Gastroenterology Hepatology (SIGEP); Italian Study Group for Turner Syndrome (ISGTS). Prevalence and clinical picture of celiac disease in Turner syndrome. J Clin Endocrinol Metab. 2002;87(12): 5495-5498.

15. Giannotti A, Tiberio G, Castro M, et al. Coeliac disease in Williams syndrome. J Med Genet. 2001;38(11):767–768.

16. Alaedini A, Green P. Narrative review: celiac disease: understanding a complex autoimmune disorder. Ann Intern Med. 2005;142(4): 289-298.

17. Fröhlich-Reiterer EE, Hofer S, Kaspers S, et al. Screening frequency for celiac disease and autoimmune thyroiditis in children and adolescents with type 1 diabetes mellitus: data from a German/Austrian multicentre survey. Pediatr Diabetes. 2008;9(6):546-553.

18. American Diabetes Association. Standards of medical care in diabetes—2009. Diabetes Care. 2009;32(1):S13–S61.

19. Losowsky MS. A history of coeliac disease. Dig Dis. 2008;26(2):112-120.

20. Evans KE, Hadjivassilou M, Sanders DS. Understanding ‘silent’ coeliac disease: complications in diagnosis and treatment. Gastrointest Nurs. 2010;8(2):26-32.

21. Lurie Y, Landau DA, Pfeffer J, Oren R. Celiac disease diagnosed in the elderly. J Clin Gastroenterol. 2008;42(1):59-61.

22. Vilppula A, Kaukinen K, Luostarinen L, et al. Increasing prevalence and high incidence of celiac disease in elderly people: a population-based study. BMC Gastroenterol. 2009 Jun 29;9:49. 

23. Lo W, Sano K, Lebwohl B, et al. Changing presentation of adult celiac disease. Dig Dis Sci. 2003;48(2):395-398.

24. Alaedini A, Okamoto H, Briani C, et al. Immune cross-reactivity in celiac disease: anti-gliadin antibodies bind to neuronal synapsin I. J Immunol. 2007;178(10):6590-6595.

25. Tursi A, Giorgetti G, Brandimarte G, et al. Prevalence and clinical presentation of subclinical/silent celiac disease in adults: an analysis on a 12-year observation. Hepatogastroenterology. 2001;48(38):462-464.

26. Rostom A, Murray JA, Kagnoff MF. American Gastroenterological Association (AGA) Institute technical review on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1981-2002.

27. Ferguson A, Arranz E, O’Mahony S. Clinical and pathological spectrum of coeliac disease—active, silent, latent, potential. Gut. 1993;34(2): 150-151.

28. Logan RFA. Problems and pitfalls in epidemiological studies of coeliac disease. In: Auricchio S, Visakorpi JK, eds. Common Food Intolerances 1. Epidemiology of Coeliac Disease (Dynamic Nutrition Research)(Pt 1). Basel, Switzerland: Karger; 1992:14-22.

29. Marsh MN. Gluten, major histocompatibility complex, and the small intestine: a molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology. 1992;102(1):330-354.

30. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol. 1999;11(10):1185-1194.

31. Corazza GR, Villanaci V. Coeliac disease.
J Clin Pathol. 2005;58(6):573-574.

32. Hadithi M, von Blomberg BM, Crusius JB, et al. Accuracy of serologic tests and HLA-DQ typing for diagnosing celiac disease. Ann Intern Med. 2007;147(5):294-302.

33. Lerner A. New therapeutic strategies for celiac disease. Autoimmun Rev. 2010;9(3):144-147.

34. Catassi C, Fasano A. Celiac disease diagnosis: simple rules are better than complicated algorithms. Am J Med. 2010;123(8):691-693.

35. Rubio-Tapia A, Kyle RA, Kaplan EL, et al. Increased prevalence and mortality in undiagnosed celiac disease. Gastroenterology. 2009; 137(1):88–93.

36. Harewood GC, Holub JL, Lieberman DA. Variation in small bowel biopsy performance among diverse endoscopy settings: results from a national endoscopic database. Am J Gastroenterol. 2004;99(9):1790-1794.

37. Brar P, Lee S, Lewis S, et al. Budesonide in the treatment of refractory celiac disease. Am J Gastroenterol. 2007;102(10):2265-2269.

38. Al-Toma A, Verbeek WHM, Hadithi M, et al. Survival in refractory coeliac disease and enteropathy-associated T-cell lymphoma: retrospective evaluation of single-centre experience. Gut. 2007;56(10):1373-1378.

39. Kaukinen K, Mäki M, Partanen J, et al. Celiac disease without villous atrophy: revision of criteria called for. Dig Dis Sci. 2001;46(4):879-887.

40. Mohamed BM, Feighery C, Coates C, et al. The absence of a mucosal lesion on standard histological examination does not exclude diagnosis of celiac disease. Dig Dis Sci. 2008; 53(1):52-61.

41. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology. 2001; 120(3):636-651.

42. US Food and Drug Administration. Topic-specific labeling information (2010). www.fda.gov/Food/LabelingNutrition/FoodLabeling GuidanceRegulatoryInformation/Topic-Specific LabelingInformation/default.htm. Accessed March 28, 2011.

43. Lindfors K, Blomqvist T, Juuti-Uusitalo K, et al. Live probiotic Bifidobacterium lactis bacteria inhibit the toxic effects induced by wheat gliadin in epithelial cell culture. Clin Exp Immunol. 2008;152(3):552-558.

44. Cellier C, Delabesse E, Helmer C, et al; French Coeliac Disease Study Group. Refractory sprue, coeliac disease, and enteropathy-associated T-cell lymphoma. Lancet. 2000;356(9225): 203-208.

45. United European Gastroenterology. When is a coeliac a coeliac? Report of a working group of the United European Gastroenterology Week in Amsterdam, 2001. Eur J Gastroenterol Hepatol. 2001;13(9):1123-1128.

46. Malamut G, Afchain P, Verkarre V, et al. Presentation and long-term follow-up of refractory celiac disease: comparison of type I with type II. Gastroenterology. 2009;136(1):81-90.

47. Al-Toma A, Goerres MS, Meijer JW, et al. Human leukocyte antigen-DQ2 homozygosity and the development of refractory celiac disease and enteropathy-associated T-cell lymphoma. Clin Gastroenterol Hepatol. 2006;4(3): 315-319.

48. Mauriño E, Niveloni S, Cherñavsky A, et al. Azathioprine in refractory sprue: results from a prospective, open-label study. Am J Gastroenterol. 2002;97(10):2595–2602.

49. Tack GJ, Verbeek WHM, Al-Toma A, et al. Evaluation of cladribine treatment in refractory celiac disease type II. World J Gastroenterol. 2011;17(4):506–513.

50. Akobeng AK, Ramanan AV, Buchan I, Heller RF. Effect of breast feeding on risk of coeliac disease: a systematic review and meta-analysis of observational studies. Arch Dis Child. 2006; 91(1):39-43.

51. Rashid M, Butzner JD, Warren R, et al. Home blood testing for celiac disease: recommendations for management. Can Fam Physician. 2009;55(2):151-153.

52. Hershcovici T, Leshno M, Goldin E, et al. Cost effectiveness of mass screening for coeliac disease is determined by time-delay to diagnosis and quality of life on a gluten-free diet. Aliment Pharmacol Ther. 2010;31(8):901-910.

Celiac disease (CD), also known as gluten-sensitive enteropathy or celiac sprue, is an endocrine disorder whose effects are triggered by the ingestion of gluten—the principle storage protein in wheat, rye, and barley.^1-3 CD inflicts damage to the mucosa of the small intestine and subsequently to systemic organ tissues. CD can affect any organ in the body.¹ The responsible genetic factors are the human leukocyte antigens, HLA -DQ2 and -DQ8, which are present in 40% of the general population but are found in nearly 100% of patients with CD.^1,3,4

Though previously considered uncommon, CD has been estimated to affect more than 1% of the general population worldwide.^1,4,5 Currently, CD is most reliably identified by ­positive serum antibodies, specifically immunoglobulin A (IgA) anti-tissue transglutaminase (tTG) and IgA antiendomysial (EMA) antibodies,⁶ and by a finding of villous atrophy of the intestinal lining on biopsy. The spectrum of presentations of CD is broad, including the “typical” intestinal features of diarrhea, bloating, abdominal pain, and weight loss; or common “atypical” extraintestinal manifestations, such as anemia, osteoporosis, infertility, and neurologic disturbances (eg, peripheral neuropathy⁷).⁸ See Table 1.^8,9

Prevalence of CD is greater among those with a family history of CD; with autoimmune diseases, especially type 1 diabetes mellitus (T1DM) and thyroiditis; and with certain genetic disorders (ie, Down, Turner, and Williams syndromes).^8-15 Because atypical features dominate in older children and adults, many cases escape diagnosis, and patients may be exposed to serious long-term complications, such as infertility and cancer.¹

CD is a lifelong condition, necessitating the complete exclusion of gluten-containing products from the diet. In the US food industry, gluten is used in numerous food applications, complicating the patient education and lifestyle changes needed to implement and maintain a gluten-free diet (GFD). However, if a GFD is not strictly followed, the patient’s quality of life can be seriously impaired.^1,4,5

AWARENESS ESSENTIAL IN PRIMARY CARE

For the primary care provider (PCP), there is no shortage of patients with gastrointestinal (GI) disorders, thyroid disease, diabetes, anemia, fatigue, or dysmenorrhea; additionally, PCPs regularly treat patients for a number of associated disorders, including anxiety, irritability, and attention deficit. Yet how likely are PCPs to screen patients with these symptoms for CD? And how many patients with CD never receive a diagnosis of the disorder?

In fact, it has been estimated that more than 90% of persons affected by CD are currently undiagnosed.^1,4 In one study involving mass screening of 1,000 children ages 2 to 18, it was determined that almost 90% of celiac-positive children had not previously been diagnosed.⁵ Similarly, in a cohort of 976 adults (median age, 54.3), the diagnostic rate for CD was initially low at 0.27 cases per thousand visits but increased to 11.6 cases per thousand visits after implementation of active screening.⁴ Based on these data, it has been estimated that more than 2.7 million Americans unknowingly carry this potentially life-threatening genetic disease.^1,4,16  

Given the potential patient population with undetected, untreated CD, some researchers consider the disorder one of the most common lifelong diseases in the US.^1,8,16 CD is closely associated with T1DM and autoimmune thyroiditis, with cross- prevalence at 11% and 6.7%, respectively.^8,12,17 The close association between T1DM and CD led the American Diabetes Association¹⁸ to amend guidelines in 2009, suggesting screening for CD in all patients newly diagnosed with T1DM.

PATIENT PRESENTATION: ADULTS VERSUS CHILDREN

Most infants and young children with CD present with the typical or “classic” triad of signs: short stature, failure to thrive, and diarrhea; in individual patients, however, the impact of genetics and exposure to gluten over time can cause considerable variation in patient presentation. As patients with undiagnosed CD age, they may present quite differently or even revert to a latent stage and become asymp­tomatic.^19,20

In two separate reviews, it was noted that classic symptoms of CD are not evident in a majority of older children and adults; instead, anemia and fatigue were the predominating symptoms.^12,20An important note: The patient with no symptoms or atypical signs of CD may still be experiencing significant damage, inflicted by gluten-induced antibodies, to the intestinal lining and/or mucosal linings in other organ systems—perhaps for years before the disease becomes evident.²⁰ 

Clinical Findings Differ With Age, Gender

Historically, CD was considered a pediatric syndrome; however, a diagnosis of CD has become increasingly common among older children and adults, especially elderly patients, although symptoms in the latter group are subtle.^21-23 Recent, active CD is being diagnosed among men older than 55 more commonly than in women of this age-group; women are generally younger at diagnosis but have experienced symptoms longer.^22,23 This later onset in men suggests that antibody seropositivity and the associated active disease may be triggered later in life.^8,22

A variety of findings have been reported in the history and physical exam of most patients who present with CD.The most prevalent signs and symptoms are abdominal pain, frequent loose stools, weight loss, joint pain, and weakness.^8,11,16 Unlike the pediatric patient with the classic triad of symptoms, adults usually experience more generalized GI manifestations, such as irritable bowel syndrome (IBS), abdominal pain, or acid reflux.¹⁰

Many patients have no GI symptoms but may present solely with fatigue, arthralgias, or myalgia.²⁰ In fact, more than 50% of adults with CD present with atypical or extraintestinal disorders, such as anemia, infertility, osteoporosis, neurologic problems, or other autoimmune disorders.^8,16,23,24 It is important for clinicians to note that atypical is somewhat typical in the older patient who presents with CD.

Patients with asymptomatic or silent CD, (see “Classification and Pathology,” below) lack both classic and atypical symptoms but still have villous atrophy, usually discovered during endoscopy being conducted for other reasons.⁸ Because of its predominantly atypical presentations, CD is considered a multisystem endocrine condition rather than one that is mainly gastrointestinal.^8,16,25,26

CLASSIFICATION AND PATHOLOGY

Though frequently a silent disorder, CD typically progresses through four stages: classical, atypical, latent, and silent. Clinicians should strive to become fully aware of each stage and its implications.^8,26,27

The classical form is primarily diagnosed in children ages 6 to 18 months. It is characterized by villous atrophy and typical symptoms of intestinal malabsorption.⁸

The patient with atypical CD has minor intestinal symptoms, but architectural abnormalities can be found in the mucosa of the small intestine. This patient is likely to present with various extraintestinal disorders, including osteoporosis, anemia, infertility, and neuropathies.^7,8

In the latent form of CD, the HLA-DQ2 and/or -DQ8 genetic markers are present. Serology for CD may be positive, but the intestinal mucosa may be normal. The patient may or may not be experiencing extraintestinal symptoms. In patients with latent CD, the gluten-associated changes appear later in life.⁸ The precise trigger for late activation of the disease, though apparently linked to genetics and gluten exposure, remains elusive.^20,24

The silent form of CD is marked by mucosal abnormalities in the small intestine and usually by positive CD serology, but it is asymptomatic. The iceberg theory of celiac disease²⁸ (see figure²⁸) has been proposed to explain CD’s hidden manifestations over time.

In patients with atypical, latent, or silent CD, the condition is sometimes detected incidentally during screening of at-risk groups or by endoscopy performed for other reasons.⁸ Most of these patients respond well to GFD therapy, noting both physical and psychological improvement—suggesting that these patients, even though asymptomatic and seemingly healthy, may have been experiencing minor manifestations of undiagnosed CD for many years: decreased appetite, fatigue, and even behavioral abnormalities.^1,8

Histopathologic analysis of abnormalities found on biopsy of the small intestine relies on the four-stage Marsh classification²⁹:

• Marsh 0: normal mucosa

• Marsh I: intraepithelial lymphocytosis

• Marsh II: intraepithelial lymphocytosis with crypt hyperplasia

• Marsh III: intraepithelial lymphocytosis with crypt hyperplasia and villous atrophy.^8,29 Modifications to this classification have been made by Oberhuber^30,31 to denote the degree of villous flattening³² (ie, IIIa, IIIb, IIIc).

Villous atrophy of the mucosa has long been considered the hallmark of CD, and its detection, according to the American Gastroenterological Association,^2,26 remains the gold standard in confirming a diagnosis of CD.^4,16,26 However, early screening (ie, serologic testing for tTG and EMA) is the necessary initial step in ensuring diagnostic accuracy, as other conditions can cause villous atrophy, and latent CD can coexist with normal intestinal mucosa.¹⁰

Avoiding Diagnostic Delays

Because of the broad spectrum of unrelated GI signs in all ages and the subtle presentation in adults, diagnosis of CD in this patient population is frequently delayed for estimated periods ranging from five to 11 years.^4,11,23,33

Improving clinician awareness of the manifestations of CD is essential³⁴; too frequently, the common symptoms of probable CD are treated as individual idiopathic disorders by both PCPs and secondary specialists, who prescribe proton pump inhibitors, antihistamines, cathartics, and/or antimotility drugs for years without ruling out a common, easily identified genetic disease. Even though the prevalence of CD has recently been shown to have increased more than fourfold since 1950,³⁵ serologic testing for CD is not widely implemented by PCPs.^4,11,20

Specialists, too, may be slow to recognize this treatable autoimmune disorder. In a recent nationwide study, it was found that gastroenterologists performed a small-bowel biopsy in less than 10% of their patients who underwent esophagogastroduodenoscopy (EGD) for likely symptoms of CD.³⁶ Relying solely on clinical expertise and visual recognition of intestinal abnormalities can delay diagnosis for years.^4,36 Many patients may never be given a correct diagnosis of CD.

The Role of Serologic Testing

Current data demonstrate that autoimmune diseases are on the rise,^8,16,36 and CD can be the primary cause or a contributing factor in several other disorders (see Table 2^8,16,37,38). Gastroenterologists may be correct in stating that biopsy is the only way to make a diagnosis of CD or to stage CD-associated intestinal damage^4,26; yet by implementing a protocol of serologic testing for tTG and EMA in at-risk patients, PCPs could prevent a missed diagnosis on EGD when biopsy has not been considered, as in the case of atypical CD; or when biopsy results are negative in a patient with latent CD.^39,40

Because of its high negative predictive value, serologic testing should be conducted first to significantly reduce the probability of suspected CD. Such selective screening should be performed by the PCP before invasive testing by the gastroenterologist and before long-term empiric treatment for idiopathic GERD, IBS, or other unexplained disorders.^32,40

Thus, it has been recommended that PCPs perform screening for CD in patients with unexplained chronic GI disturbances or a familial prevalence of CD, or in those who present with the atypical signs of CD or with associated disorders.^1,10,16,20 Whether serologic screening results are positive or negative for CD, the patient with classic GI symptoms should undergo endoscopy with biopsy to confirm active disease and to evaluate the extent of intestinal damage—or to explore other causes.^26,39 An algorithm^{2,4,8,11,16,26} illustrating suggested screening, treatment, and follow-up strategies for patients at high risk for CD is shown below.

Catassi and Fasano³⁴ recently proposed a “four out of five” rule, by which diagnosis of CD may be confirmed in patients with at least four of the following five criteria:

• Typical symptoms of CD

• Positive serology (ie, IgA tTG and IgA EMA antibodies)

• Genetic susceptibility (as confirmed by the presence of HLA-DQ2 and HLA-DQ8)

• Small intestine biopsy results indicating celiac enteropathy

• Improvement of CD signs and symptoms following implementation of the GFD.³⁴

CURRENT TREATMENT AND ASSOCIATED CHALLENGES

Because gluten consumption is the principal trigger of CD pathology, a GFD is considered the safest, most effective therapy for the disorder.^{1,8,10,11,16,19} Implementing and maintaining the GFD involves a considerable learning curve for the patient, the patient’s family, and possibly the provider; to achieve complete recovery, all involved must become knowledgeable regarding gluten-free and gluten-containing products. The patient must be willing and able to avoid those that contain gluten and bear the potentially high costs⁸ of gluten-free foods.

Even for patients with CD who are determined to comply with the GFD, gluten monitoring can be difficult. There are ways to determine what is a safe level of gluten ingestion for each patient, but trace amounts of gluten are found in many products, including some that are marked “gluten-free.”^1,41 The FDA has proposed that a product labeled gluten-free may contain no more than 20 parts per million (ppm, ie, 20 mg/kg) of gluten.⁴² In other countries, however, acceptable levels may be as high as 200 ppm (200 mg/kg)—which are considered well above the trigger amounts in the average patient with CD.^1,41 The complex nature of each patient’s sensitivity to gluten and the ubiquitous presence of gluten as a food source in both industrialized and developing countries make adherence to the GFD challenging.¹⁰

It is critical for the PCP to help the patient review all of his or her prescription and OTC pharmaceuticals and nutritional supplements, as these may contain hidden gluten in the form of modified starches and other fillers.⁴¹ It may be also advisable to involve the patient’s pharmacist, requesting an assessment for agents that may be suspect.

A management team approach may ensure the most integrative care. In addition to the PCP and the pharmacist, such a team might include a gastroenterologist, an endocrinologist, a nutritionist, and a psychologist, who may be needed to help the patient confront the great life adjustment required, in addition to addressing other behavioral disorders that are common in patients with CD.^10,26

See the box for resources that may be beneficial for both patients and their clinicians.

Alternative Medicine Options

Alternative medicine is gaining favor, especially when no drug therapy is currently available to alleviate gluten toxicity. Supplementation with the fat-soluble vitamins (A, D, E, and K), vitamin B₁₂, folic acid, and the minerals calcium and iron, as indicated by serum deficiencies, is recommended.^10,20 Supplementation with digestive enzymes, which are known to be deficient in patients with CD as a result of villous atrophy, may help break down undigested gluten proteins; research is under way to find a recombinant enzyme therapy.¹⁰ Researchers have recently shown that probiotics (specifically, Bifidobacterium lactis) significantly reduce the immune response when incidental exposure to gluten occurs.⁴³ 

REFRACTORY CELIAC DISEASE

Patients with late-onset CD, especially those not diagnosed until after age 50, may have a diminished or absent response to dietary therapy. In some patients, histologic signs and clinical symptoms persist or relapse after a prior positive response to a strict GFD, despite continued adherence to the diet for longer than 12 months.⁴⁴ Once other causes have been carefully excluded, these patients are considered to have refractory celiac disease (RCD). Exact prevalence of RCD is unknown, but Tack et al⁸ estimate it at 5% of all cases of CD. Relapsing CD resulting from poor adherence to the GFD is not considered true RCD.^8,16,37,38

According to researchers for the European Celiac Disease working group,^8,45 RCD can be divided into types I and II:

• RCD I, in which normal polyclonal T cells are present in the intestinal lumen

• RCD II, in which abnormal clonal T cells infiltrate the intestinal mucosa, representing premalignancy.⁴⁵

The histologic picture of RCD mimics that of severe CD. Malabsorption complications, lesions in the intestinal mucosa, and inflammatory lymphocytosis are present.⁴⁴ Some patients, like those with classical CD, have serology test results that are consistent with CD and an initial response to GFD therapy; after months or years, however, this response subsides. Other patients are immediately unresponsive to GFD and lack the serologic markers for CD.⁸

A differential diagnosis including other explanations for the manifestations of RCD must be carefully reviewed, with each excluded, through the strategies shown in Table 2. This review is essential, as patients with RCD II have a much worse prognosis than those with RCD I; the associated five-year survival rates are 44% to 58%, versus 85% or greater, respectively.^36,46

Additionally, the continued autoimmune expansion of aberrant T cells in patients with RCD II causes early conversion to malignancy, usually within four to six years after diagnosis. Enteropathy-associated T-cell lymphoma is the most common malignancy, occurring in more than 50% of patients with RCD II, and a likely cause of death.^3,8,46,47

Treatment for Refractory Celiac Disease

In addition to the GFD, patients with RCD I generally respond well to corticosteroids or other immunosuppressive treatment.⁸ Use of budesonide, a corticosteroid given in a once-daily, 9-mg dose, has led to almost complete recovery in most patients. Duration of therapy is response-dependent.³⁷

Systemic corticosteroids or other immunosuppressant agents, such as azathioprine, should be reserved for patients with RCD I or RCD II who do not respond to budesonide, as lengthy treatment regimens are required, with considerable risk for adverse effects.^35,48

Recently, promising results have been reported in a small, open-label cohort study involving patients with RCD II who underwent five days of treatment with IV cladribine (0.1 mg/kg/d).⁴⁹

PREVENTION OF CD

A good nutritional start from birth could be the best means of preventing symptomatic CD. According to findings from a meta-analysis of data from four studies, children being breastfed at the time gluten was introduced had a 52% reduction in risk for CD, compared with their peers who were not being breastfed at that time.⁵⁰

The protection breast milk appears to provide against CD is not clearly understood. One possible mechanism is that breast milk may protect an infant against CD by preventing gastrointestinal infections, as is the case with other infections. The presence of GI infections (eg, rotavirus) in early life could lead to increased permeability of the intestinal mucosa, allowing the passage of gluten into the lamina propria.^3,8,50

Extended duration of breastfeeding is also associated with a reduced risk for CD.^8,41,50 Long-term studies are needed, however, to determine whether breastfeeding delays CD onset or provides permanent protection against the disorder.

RECENT DEVELOPMENTS

A recently marketed OTC testing kit for CD is now available in Canada and other countries outside the US; this may be an indication of the growing awareness of the numbers of patients with undiagnosed CD. The test parallels the tTG serum test, which in the US is evaluated only in laboratories; it has comparable specificity and sensitivity, with results within 10 minutes. In the US, the FDA has not yet approved the kit, but domestic testing of the product may soon be under way.⁵¹

Alternative treatment modalities are currently focusing on the detoxification of wheat components, rapid enzymatic degradation to reduce exposure to gluten, inducing gluten tolerance, inhibiting permeability of the small intestine to gluten (which, it is thought, may prevent many of the systemic manifestations of CD), and finally, development of an immunomodulatory vaccine.^8,33 None of these therapies is yet approved.

IMPLICATIONS OF DELAYED DIAGNOSIS

The unrecognized prevalence of CD is a growing issue, as many symptomatic but unscreened patients are frequently misdiagnosed with IBS, chronic fatigue, or other idiopathic disorders. The silent and latent forms of CD are of the greatest concern, as they show minimal signs and can lead to multiple organ system damage and are implicated in other autoimmune disorders. The longer diagnosis is delayed, the greater is patients’ resistance to dietary therapy, and the less likely that established intestinal and/or neurologic damage can be reversed.^10,20,51

The large proportion of undiagnosed celiac patients may account for an accompanying underestimated cost to both the patient and the health care system because of repeated referrals to investigate unexplained disorders before an accurate diagnosis is made. In one recent analysis, mass screening for CD in a young adult population led to improved quality-of-life years by shortening the time to diagnosis and treatment; it was also found cost-effective.⁵² PCPs must be attentive to patients who may be at high risk for CD and implement combined serum tTG and EMA screening as the initial step in identification and treatment.^4,10,11,20 Some form of standardized screening protocol may become inevitable.

CONCLUSION

The prevalence of CD has increased more than fourfold since 1950, and diagnosis is often significantly delayed. Increased awareness is needed among PCPs that CD in adults is likely to manifest with atypical (ie, nongastrointestinal) symptoms and signs. Judicious use of serologic screening for CD would lead to earlier diagnosis and more effective treatment, possibly preventing the potentially lethal refractory disease forms associated with chronic untreated CD.  

REFERENCES

1. Ramos M, Orozovich P, Moser K, et al. Health 1. Catassi C, Fasano A. Celiac disease. Curr Opin Gastroenterol. 2008;24(6):687-691.

2. AGA Institute. AGA Institute medical position statement on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1977-1980.

3. Green PH, Jabri B. Coeliac disease. Lancet. 2003;362(9381):383-391.

4. Catassi C, Kryszak D, Louis-Jacques O, et al. Detection of celiac disease in primary care: a multicenter case-finding study in North America. Am J Gastroenterol. 2007;102(7);1454-1460.

5. Demirçeken FG, Kansu A, Kuloglu Z, et al. Human tissue transglutaminase antibody screening by immunochromatographic line immunoassay for early diagnosis of celiac disease in Turkish children. Turk J Gastroenterol. 2008;19(1):14-21.

6. van der Windt DA, Jellema P, Mulder CJ, et al. Diagnostic testing for celiac disease among patients with abdominal symptoms: a systematic review. JAMA. 2010;303(17):1738-1746.

7. Freeman HJ. Neurological disorders in adult celiac disease. Can J Gastroenterol. 2008; 22(11):909-911.

8. Tack GJ, Verbeek WHM, Schreurs MWJ, Mulder CJJ. The spectrum of celiac disease: epidemiology, clinical aspects and treatment. Nat Rev Gastroenterol Hepatol. 2010;7(4):204-213.

9. Farrell RJ, Kelly CP. Celiac sprue. N Engl J Med. 2002;346(3):180-188.

10. Green PHR, Cellier C. Celiac disease. N Engl J Med. 2007;357(17):1731-1743.

11. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003;163(3):286-292.

12. Sud S, Marcon M, Assor E, et al. Celiac disease and pediatric type 1 diabetes: diagnostic and treatment dilemmas. Int J Pediatr Endocrinol. 2010;2010:161285. Epub 2010 Jun 23.

13. Swigonski NL, Kuhlenschmidt HL, Bull MJ, et al. Screening for celiac disease in asymptomatic children with Down syndrome: cost-effectiveness of preventing lymphoma. Pediatrics. 2006;118(2):594-602.

14. Bonamico M, Pasquino AM, Mariani P, et al; Italian Society of Pediatric Gastroenterology Hepatology (SIGEP); Italian Study Group for Turner Syndrome (ISGTS). Prevalence and clinical picture of celiac disease in Turner syndrome. J Clin Endocrinol Metab. 2002;87(12): 5495-5498.

15. Giannotti A, Tiberio G, Castro M, et al. Coeliac disease in Williams syndrome. J Med Genet. 2001;38(11):767–768.

16. Alaedini A, Green P. Narrative review: celiac disease: understanding a complex autoimmune disorder. Ann Intern Med. 2005;142(4): 289-298.

17. Fröhlich-Reiterer EE, Hofer S, Kaspers S, et al. Screening frequency for celiac disease and autoimmune thyroiditis in children and adolescents with type 1 diabetes mellitus: data from a German/Austrian multicentre survey. Pediatr Diabetes. 2008;9(6):546-553.

18. American Diabetes Association. Standards of medical care in diabetes—2009. Diabetes Care. 2009;32(1):S13–S61.

19. Losowsky MS. A history of coeliac disease. Dig Dis. 2008;26(2):112-120.

20. Evans KE, Hadjivassilou M, Sanders DS. Understanding ‘silent’ coeliac disease: complications in diagnosis and treatment. Gastrointest Nurs. 2010;8(2):26-32.

21. Lurie Y, Landau DA, Pfeffer J, Oren R. Celiac disease diagnosed in the elderly. J Clin Gastroenterol. 2008;42(1):59-61.

22. Vilppula A, Kaukinen K, Luostarinen L, et al. Increasing prevalence and high incidence of celiac disease in elderly people: a population-based study. BMC Gastroenterol. 2009 Jun 29;9:49. 

23. Lo W, Sano K, Lebwohl B, et al. Changing presentation of adult celiac disease. Dig Dis Sci. 2003;48(2):395-398.

24. Alaedini A, Okamoto H, Briani C, et al. Immune cross-reactivity in celiac disease: anti-gliadin antibodies bind to neuronal synapsin I. J Immunol. 2007;178(10):6590-6595.

25. Tursi A, Giorgetti G, Brandimarte G, et al. Prevalence and clinical presentation of subclinical/silent celiac disease in adults: an analysis on a 12-year observation. Hepatogastroenterology. 2001;48(38):462-464.

26. Rostom A, Murray JA, Kagnoff MF. American Gastroenterological Association (AGA) Institute technical review on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1981-2002.

27. Ferguson A, Arranz E, O’Mahony S. Clinical and pathological spectrum of coeliac disease—active, silent, latent, potential. Gut. 1993;34(2): 150-151.

28. Logan RFA. Problems and pitfalls in epidemiological studies of coeliac disease. In: Auricchio S, Visakorpi JK, eds. Common Food Intolerances 1. Epidemiology of Coeliac Disease (Dynamic Nutrition Research)(Pt 1). Basel, Switzerland: Karger; 1992:14-22.

29. Marsh MN. Gluten, major histocompatibility complex, and the small intestine: a molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology. 1992;102(1):330-354.

30. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol. 1999;11(10):1185-1194.

31. Corazza GR, Villanaci V. Coeliac disease.
J Clin Pathol. 2005;58(6):573-574.

32. Hadithi M, von Blomberg BM, Crusius JB, et al. Accuracy of serologic tests and HLA-DQ typing for diagnosing celiac disease. Ann Intern Med. 2007;147(5):294-302.

33. Lerner A. New therapeutic strategies for celiac disease. Autoimmun Rev. 2010;9(3):144-147.

34. Catassi C, Fasano A. Celiac disease diagnosis: simple rules are better than complicated algorithms. Am J Med. 2010;123(8):691-693.

35. Rubio-Tapia A, Kyle RA, Kaplan EL, et al. Increased prevalence and mortality in undiagnosed celiac disease. Gastroenterology. 2009; 137(1):88–93.

36. Harewood GC, Holub JL, Lieberman DA. Variation in small bowel biopsy performance among diverse endoscopy settings: results from a national endoscopic database. Am J Gastroenterol. 2004;99(9):1790-1794.

37. Brar P, Lee S, Lewis S, et al. Budesonide in the treatment of refractory celiac disease. Am J Gastroenterol. 2007;102(10):2265-2269.

38. Al-Toma A, Verbeek WHM, Hadithi M, et al. Survival in refractory coeliac disease and enteropathy-associated T-cell lymphoma: retrospective evaluation of single-centre experience. Gut. 2007;56(10):1373-1378.

39. Kaukinen K, Mäki M, Partanen J, et al. Celiac disease without villous atrophy: revision of criteria called for. Dig Dis Sci. 2001;46(4):879-887.

40. Mohamed BM, Feighery C, Coates C, et al. The absence of a mucosal lesion on standard histological examination does not exclude diagnosis of celiac disease. Dig Dis Sci. 2008; 53(1):52-61.

41. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology. 2001; 120(3):636-651.

42. US Food and Drug Administration. Topic-specific labeling information (2010). www.fda.gov/Food/LabelingNutrition/FoodLabeling GuidanceRegulatoryInformation/Topic-Specific LabelingInformation/default.htm. Accessed March 28, 2011.

43. Lindfors K, Blomqvist T, Juuti-Uusitalo K, et al. Live probiotic Bifidobacterium lactis bacteria inhibit the toxic effects induced by wheat gliadin in epithelial cell culture. Clin Exp Immunol. 2008;152(3):552-558.

44. Cellier C, Delabesse E, Helmer C, et al; French Coeliac Disease Study Group. Refractory sprue, coeliac disease, and enteropathy-associated T-cell lymphoma. Lancet. 2000;356(9225): 203-208.

45. United European Gastroenterology. When is a coeliac a coeliac? Report of a working group of the United European Gastroenterology Week in Amsterdam, 2001. Eur J Gastroenterol Hepatol. 2001;13(9):1123-1128.

46. Malamut G, Afchain P, Verkarre V, et al. Presentation and long-term follow-up of refractory celiac disease: comparison of type I with type II. Gastroenterology. 2009;136(1):81-90.

47. Al-Toma A, Goerres MS, Meijer JW, et al. Human leukocyte antigen-DQ2 homozygosity and the development of refractory celiac disease and enteropathy-associated T-cell lymphoma. Clin Gastroenterol Hepatol. 2006;4(3): 315-319.

48. Mauriño E, Niveloni S, Cherñavsky A, et al. Azathioprine in refractory sprue: results from a prospective, open-label study. Am J Gastroenterol. 2002;97(10):2595–2602.

49. Tack GJ, Verbeek WHM, Al-Toma A, et al. Evaluation of cladribine treatment in refractory celiac disease type II. World J Gastroenterol. 2011;17(4):506–513.

50. Akobeng AK, Ramanan AV, Buchan I, Heller RF. Effect of breast feeding on risk of coeliac disease: a systematic review and meta-analysis of observational studies. Arch Dis Child. 2006; 91(1):39-43.

51. Rashid M, Butzner JD, Warren R, et al. Home blood testing for celiac disease: recommendations for management. Can Fam Physician. 2009;55(2):151-153.

52. Hershcovici T, Leshno M, Goldin E, et al. Cost effectiveness of mass screening for coeliac disease is determined by time-delay to diagnosis and quality of life on a gluten-free diet. Aliment Pharmacol Ther. 2010;31(8):901-910.

Celiac disease (CD), also known as gluten-sensitive enteropathy or celiac sprue, is an endocrine disorder whose effects are triggered by the ingestion of gluten—the principle storage protein in wheat, rye, and barley.^1-3 CD inflicts damage to the mucosa of the small intestine and subsequently to systemic organ tissues. CD can affect any organ in the body.¹ The responsible genetic factors are the human leukocyte antigens, HLA -DQ2 and -DQ8, which are present in 40% of the general population but are found in nearly 100% of patients with CD.^1,3,4

Though previously considered uncommon, CD has been estimated to affect more than 1% of the general population worldwide.^1,4,5 Currently, CD is most reliably identified by ­positive serum antibodies, specifically immunoglobulin A (IgA) anti-tissue transglutaminase (tTG) and IgA antiendomysial (EMA) antibodies,⁶ and by a finding of villous atrophy of the intestinal lining on biopsy. The spectrum of presentations of CD is broad, including the “typical” intestinal features of diarrhea, bloating, abdominal pain, and weight loss; or common “atypical” extraintestinal manifestations, such as anemia, osteoporosis, infertility, and neurologic disturbances (eg, peripheral neuropathy⁷).⁸ See Table 1.^8,9

Prevalence of CD is greater among those with a family history of CD; with autoimmune diseases, especially type 1 diabetes mellitus (T1DM) and thyroiditis; and with certain genetic disorders (ie, Down, Turner, and Williams syndromes).^8-15 Because atypical features dominate in older children and adults, many cases escape diagnosis, and patients may be exposed to serious long-term complications, such as infertility and cancer.¹

CD is a lifelong condition, necessitating the complete exclusion of gluten-containing products from the diet. In the US food industry, gluten is used in numerous food applications, complicating the patient education and lifestyle changes needed to implement and maintain a gluten-free diet (GFD). However, if a GFD is not strictly followed, the patient’s quality of life can be seriously impaired.^1,4,5

AWARENESS ESSENTIAL IN PRIMARY CARE

For the primary care provider (PCP), there is no shortage of patients with gastrointestinal (GI) disorders, thyroid disease, diabetes, anemia, fatigue, or dysmenorrhea; additionally, PCPs regularly treat patients for a number of associated disorders, including anxiety, irritability, and attention deficit. Yet how likely are PCPs to screen patients with these symptoms for CD? And how many patients with CD never receive a diagnosis of the disorder?

In fact, it has been estimated that more than 90% of persons affected by CD are currently undiagnosed.^1,4 In one study involving mass screening of 1,000 children ages 2 to 18, it was determined that almost 90% of celiac-positive children had not previously been diagnosed.⁵ Similarly, in a cohort of 976 adults (median age, 54.3), the diagnostic rate for CD was initially low at 0.27 cases per thousand visits but increased to 11.6 cases per thousand visits after implementation of active screening.⁴ Based on these data, it has been estimated that more than 2.7 million Americans unknowingly carry this potentially life-threatening genetic disease.^1,4,16  

Given the potential patient population with undetected, untreated CD, some researchers consider the disorder one of the most common lifelong diseases in the US.^1,8,16 CD is closely associated with T1DM and autoimmune thyroiditis, with cross- prevalence at 11% and 6.7%, respectively.^8,12,17 The close association between T1DM and CD led the American Diabetes Association¹⁸ to amend guidelines in 2009, suggesting screening for CD in all patients newly diagnosed with T1DM.

PATIENT PRESENTATION: ADULTS VERSUS CHILDREN

Most infants and young children with CD present with the typical or “classic” triad of signs: short stature, failure to thrive, and diarrhea; in individual patients, however, the impact of genetics and exposure to gluten over time can cause considerable variation in patient presentation. As patients with undiagnosed CD age, they may present quite differently or even revert to a latent stage and become asymp­tomatic.^19,20

In two separate reviews, it was noted that classic symptoms of CD are not evident in a majority of older children and adults; instead, anemia and fatigue were the predominating symptoms.^12,20An important note: The patient with no symptoms or atypical signs of CD may still be experiencing significant damage, inflicted by gluten-induced antibodies, to the intestinal lining and/or mucosal linings in other organ systems—perhaps for years before the disease becomes evident.²⁰ 

Clinical Findings Differ With Age, Gender

Historically, CD was considered a pediatric syndrome; however, a diagnosis of CD has become increasingly common among older children and adults, especially elderly patients, although symptoms in the latter group are subtle.^21-23 Recent, active CD is being diagnosed among men older than 55 more commonly than in women of this age-group; women are generally younger at diagnosis but have experienced symptoms longer.^22,23 This later onset in men suggests that antibody seropositivity and the associated active disease may be triggered later in life.^8,22

A variety of findings have been reported in the history and physical exam of most patients who present with CD.The most prevalent signs and symptoms are abdominal pain, frequent loose stools, weight loss, joint pain, and weakness.^8,11,16 Unlike the pediatric patient with the classic triad of symptoms, adults usually experience more generalized GI manifestations, such as irritable bowel syndrome (IBS), abdominal pain, or acid reflux.¹⁰

Many patients have no GI symptoms but may present solely with fatigue, arthralgias, or myalgia.²⁰ In fact, more than 50% of adults with CD present with atypical or extraintestinal disorders, such as anemia, infertility, osteoporosis, neurologic problems, or other autoimmune disorders.^8,16,23,24 It is important for clinicians to note that atypical is somewhat typical in the older patient who presents with CD.

Patients with asymptomatic or silent CD, (see “Classification and Pathology,” below) lack both classic and atypical symptoms but still have villous atrophy, usually discovered during endoscopy being conducted for other reasons.⁸ Because of its predominantly atypical presentations, CD is considered a multisystem endocrine condition rather than one that is mainly gastrointestinal.^8,16,25,26

CLASSIFICATION AND PATHOLOGY

Though frequently a silent disorder, CD typically progresses through four stages: classical, atypical, latent, and silent. Clinicians should strive to become fully aware of each stage and its implications.^8,26,27

The classical form is primarily diagnosed in children ages 6 to 18 months. It is characterized by villous atrophy and typical symptoms of intestinal malabsorption.⁸

The patient with atypical CD has minor intestinal symptoms, but architectural abnormalities can be found in the mucosa of the small intestine. This patient is likely to present with various extraintestinal disorders, including osteoporosis, anemia, infertility, and neuropathies.^7,8

In the latent form of CD, the HLA-DQ2 and/or -DQ8 genetic markers are present. Serology for CD may be positive, but the intestinal mucosa may be normal. The patient may or may not be experiencing extraintestinal symptoms. In patients with latent CD, the gluten-associated changes appear later in life.⁸ The precise trigger for late activation of the disease, though apparently linked to genetics and gluten exposure, remains elusive.^20,24

The silent form of CD is marked by mucosal abnormalities in the small intestine and usually by positive CD serology, but it is asymptomatic. The iceberg theory of celiac disease²⁸ (see figure²⁸) has been proposed to explain CD’s hidden manifestations over time.

In patients with atypical, latent, or silent CD, the condition is sometimes detected incidentally during screening of at-risk groups or by endoscopy performed for other reasons.⁸ Most of these patients respond well to GFD therapy, noting both physical and psychological improvement—suggesting that these patients, even though asymptomatic and seemingly healthy, may have been experiencing minor manifestations of undiagnosed CD for many years: decreased appetite, fatigue, and even behavioral abnormalities.^1,8

Histopathologic analysis of abnormalities found on biopsy of the small intestine relies on the four-stage Marsh classification²⁹:

• Marsh 0: normal mucosa

• Marsh I: intraepithelial lymphocytosis

• Marsh II: intraepithelial lymphocytosis with crypt hyperplasia

• Marsh III: intraepithelial lymphocytosis with crypt hyperplasia and villous atrophy.^8,29 Modifications to this classification have been made by Oberhuber^30,31 to denote the degree of villous flattening³² (ie, IIIa, IIIb, IIIc).

Villous atrophy of the mucosa has long been considered the hallmark of CD, and its detection, according to the American Gastroenterological Association,^2,26 remains the gold standard in confirming a diagnosis of CD.^4,16,26 However, early screening (ie, serologic testing for tTG and EMA) is the necessary initial step in ensuring diagnostic accuracy, as other conditions can cause villous atrophy, and latent CD can coexist with normal intestinal mucosa.¹⁰

Avoiding Diagnostic Delays

Because of the broad spectrum of unrelated GI signs in all ages and the subtle presentation in adults, diagnosis of CD in this patient population is frequently delayed for estimated periods ranging from five to 11 years.^4,11,23,33

Improving clinician awareness of the manifestations of CD is essential³⁴; too frequently, the common symptoms of probable CD are treated as individual idiopathic disorders by both PCPs and secondary specialists, who prescribe proton pump inhibitors, antihistamines, cathartics, and/or antimotility drugs for years without ruling out a common, easily identified genetic disease. Even though the prevalence of CD has recently been shown to have increased more than fourfold since 1950,³⁵ serologic testing for CD is not widely implemented by PCPs.^4,11,20

Specialists, too, may be slow to recognize this treatable autoimmune disorder. In a recent nationwide study, it was found that gastroenterologists performed a small-bowel biopsy in less than 10% of their patients who underwent esophagogastroduodenoscopy (EGD) for likely symptoms of CD.³⁶ Relying solely on clinical expertise and visual recognition of intestinal abnormalities can delay diagnosis for years.^4,36 Many patients may never be given a correct diagnosis of CD.

The Role of Serologic Testing

Current data demonstrate that autoimmune diseases are on the rise,^8,16,36 and CD can be the primary cause or a contributing factor in several other disorders (see Table 2^8,16,37,38). Gastroenterologists may be correct in stating that biopsy is the only way to make a diagnosis of CD or to stage CD-associated intestinal damage^4,26; yet by implementing a protocol of serologic testing for tTG and EMA in at-risk patients, PCPs could prevent a missed diagnosis on EGD when biopsy has not been considered, as in the case of atypical CD; or when biopsy results are negative in a patient with latent CD.^39,40

Because of its high negative predictive value, serologic testing should be conducted first to significantly reduce the probability of suspected CD. Such selective screening should be performed by the PCP before invasive testing by the gastroenterologist and before long-term empiric treatment for idiopathic GERD, IBS, or other unexplained disorders.^32,40

Thus, it has been recommended that PCPs perform screening for CD in patients with unexplained chronic GI disturbances or a familial prevalence of CD, or in those who present with the atypical signs of CD or with associated disorders.^1,10,16,20 Whether serologic screening results are positive or negative for CD, the patient with classic GI symptoms should undergo endoscopy with biopsy to confirm active disease and to evaluate the extent of intestinal damage—or to explore other causes.^26,39 An algorithm^{2,4,8,11,16,26} illustrating suggested screening, treatment, and follow-up strategies for patients at high risk for CD is shown below.

Catassi and Fasano³⁴ recently proposed a “four out of five” rule, by which diagnosis of CD may be confirmed in patients with at least four of the following five criteria:

• Typical symptoms of CD

• Positive serology (ie, IgA tTG and IgA EMA antibodies)

• Genetic susceptibility (as confirmed by the presence of HLA-DQ2 and HLA-DQ8)

• Small intestine biopsy results indicating celiac enteropathy

• Improvement of CD signs and symptoms following implementation of the GFD.³⁴

CURRENT TREATMENT AND ASSOCIATED CHALLENGES

Because gluten consumption is the principal trigger of CD pathology, a GFD is considered the safest, most effective therapy for the disorder.^{1,8,10,11,16,19} Implementing and maintaining the GFD involves a considerable learning curve for the patient, the patient’s family, and possibly the provider; to achieve complete recovery, all involved must become knowledgeable regarding gluten-free and gluten-containing products. The patient must be willing and able to avoid those that contain gluten and bear the potentially high costs⁸ of gluten-free foods.

Even for patients with CD who are determined to comply with the GFD, gluten monitoring can be difficult. There are ways to determine what is a safe level of gluten ingestion for each patient, but trace amounts of gluten are found in many products, including some that are marked “gluten-free.”^1,41 The FDA has proposed that a product labeled gluten-free may contain no more than 20 parts per million (ppm, ie, 20 mg/kg) of gluten.⁴² In other countries, however, acceptable levels may be as high as 200 ppm (200 mg/kg)—which are considered well above the trigger amounts in the average patient with CD.^1,41 The complex nature of each patient’s sensitivity to gluten and the ubiquitous presence of gluten as a food source in both industrialized and developing countries make adherence to the GFD challenging.¹⁰

It is critical for the PCP to help the patient review all of his or her prescription and OTC pharmaceuticals and nutritional supplements, as these may contain hidden gluten in the form of modified starches and other fillers.⁴¹ It may be also advisable to involve the patient’s pharmacist, requesting an assessment for agents that may be suspect.

A management team approach may ensure the most integrative care. In addition to the PCP and the pharmacist, such a team might include a gastroenterologist, an endocrinologist, a nutritionist, and a psychologist, who may be needed to help the patient confront the great life adjustment required, in addition to addressing other behavioral disorders that are common in patients with CD.^10,26

See the box for resources that may be beneficial for both patients and their clinicians.

Alternative Medicine Options

Alternative medicine is gaining favor, especially when no drug therapy is currently available to alleviate gluten toxicity. Supplementation with the fat-soluble vitamins (A, D, E, and K), vitamin B₁₂, folic acid, and the minerals calcium and iron, as indicated by serum deficiencies, is recommended.^10,20 Supplementation with digestive enzymes, which are known to be deficient in patients with CD as a result of villous atrophy, may help break down undigested gluten proteins; research is under way to find a recombinant enzyme therapy.¹⁰ Researchers have recently shown that probiotics (specifically, Bifidobacterium lactis) significantly reduce the immune response when incidental exposure to gluten occurs.⁴³ 

REFRACTORY CELIAC DISEASE

Patients with late-onset CD, especially those not diagnosed until after age 50, may have a diminished or absent response to dietary therapy. In some patients, histologic signs and clinical symptoms persist or relapse after a prior positive response to a strict GFD, despite continued adherence to the diet for longer than 12 months.⁴⁴ Once other causes have been carefully excluded, these patients are considered to have refractory celiac disease (RCD). Exact prevalence of RCD is unknown, but Tack et al⁸ estimate it at 5% of all cases of CD. Relapsing CD resulting from poor adherence to the GFD is not considered true RCD.^8,16,37,38

According to researchers for the European Celiac Disease working group,^8,45 RCD can be divided into types I and II:

• RCD I, in which normal polyclonal T cells are present in the intestinal lumen

• RCD II, in which abnormal clonal T cells infiltrate the intestinal mucosa, representing premalignancy.⁴⁵

The histologic picture of RCD mimics that of severe CD. Malabsorption complications, lesions in the intestinal mucosa, and inflammatory lymphocytosis are present.⁴⁴ Some patients, like those with classical CD, have serology test results that are consistent with CD and an initial response to GFD therapy; after months or years, however, this response subsides. Other patients are immediately unresponsive to GFD and lack the serologic markers for CD.⁸

A differential diagnosis including other explanations for the manifestations of RCD must be carefully reviewed, with each excluded, through the strategies shown in Table 2. This review is essential, as patients with RCD II have a much worse prognosis than those with RCD I; the associated five-year survival rates are 44% to 58%, versus 85% or greater, respectively.^36,46

Additionally, the continued autoimmune expansion of aberrant T cells in patients with RCD II causes early conversion to malignancy, usually within four to six years after diagnosis. Enteropathy-associated T-cell lymphoma is the most common malignancy, occurring in more than 50% of patients with RCD II, and a likely cause of death.^3,8,46,47

Treatment for Refractory Celiac Disease

In addition to the GFD, patients with RCD I generally respond well to corticosteroids or other immunosuppressive treatment.⁸ Use of budesonide, a corticosteroid given in a once-daily, 9-mg dose, has led to almost complete recovery in most patients. Duration of therapy is response-dependent.³⁷

Systemic corticosteroids or other immunosuppressant agents, such as azathioprine, should be reserved for patients with RCD I or RCD II who do not respond to budesonide, as lengthy treatment regimens are required, with considerable risk for adverse effects.^35,48

Recently, promising results have been reported in a small, open-label cohort study involving patients with RCD II who underwent five days of treatment with IV cladribine (0.1 mg/kg/d).⁴⁹

PREVENTION OF CD

A good nutritional start from birth could be the best means of preventing symptomatic CD. According to findings from a meta-analysis of data from four studies, children being breastfed at the time gluten was introduced had a 52% reduction in risk for CD, compared with their peers who were not being breastfed at that time.⁵⁰

The protection breast milk appears to provide against CD is not clearly understood. One possible mechanism is that breast milk may protect an infant against CD by preventing gastrointestinal infections, as is the case with other infections. The presence of GI infections (eg, rotavirus) in early life could lead to increased permeability of the intestinal mucosa, allowing the passage of gluten into the lamina propria.^3,8,50

Extended duration of breastfeeding is also associated with a reduced risk for CD.^8,41,50 Long-term studies are needed, however, to determine whether breastfeeding delays CD onset or provides permanent protection against the disorder.

RECENT DEVELOPMENTS

A recently marketed OTC testing kit for CD is now available in Canada and other countries outside the US; this may be an indication of the growing awareness of the numbers of patients with undiagnosed CD. The test parallels the tTG serum test, which in the US is evaluated only in laboratories; it has comparable specificity and sensitivity, with results within 10 minutes. In the US, the FDA has not yet approved the kit, but domestic testing of the product may soon be under way.⁵¹

Alternative treatment modalities are currently focusing on the detoxification of wheat components, rapid enzymatic degradation to reduce exposure to gluten, inducing gluten tolerance, inhibiting permeability of the small intestine to gluten (which, it is thought, may prevent many of the systemic manifestations of CD), and finally, development of an immunomodulatory vaccine.^8,33 None of these therapies is yet approved.

IMPLICATIONS OF DELAYED DIAGNOSIS

The unrecognized prevalence of CD is a growing issue, as many symptomatic but unscreened patients are frequently misdiagnosed with IBS, chronic fatigue, or other idiopathic disorders. The silent and latent forms of CD are of the greatest concern, as they show minimal signs and can lead to multiple organ system damage and are implicated in other autoimmune disorders. The longer diagnosis is delayed, the greater is patients’ resistance to dietary therapy, and the less likely that established intestinal and/or neurologic damage can be reversed.^10,20,51

The large proportion of undiagnosed celiac patients may account for an accompanying underestimated cost to both the patient and the health care system because of repeated referrals to investigate unexplained disorders before an accurate diagnosis is made. In one recent analysis, mass screening for CD in a young adult population led to improved quality-of-life years by shortening the time to diagnosis and treatment; it was also found cost-effective.⁵² PCPs must be attentive to patients who may be at high risk for CD and implement combined serum tTG and EMA screening as the initial step in identification and treatment.^4,10,11,20 Some form of standardized screening protocol may become inevitable.

CONCLUSION

The prevalence of CD has increased more than fourfold since 1950, and diagnosis is often significantly delayed. Increased awareness is needed among PCPs that CD in adults is likely to manifest with atypical (ie, nongastrointestinal) symptoms and signs. Judicious use of serologic screening for CD would lead to earlier diagnosis and more effective treatment, possibly preventing the potentially lethal refractory disease forms associated with chronic untreated CD.  

REFERENCES

1. Ramos M, Orozovich P, Moser K, et al. Health 1. Catassi C, Fasano A. Celiac disease. Curr Opin Gastroenterol. 2008;24(6):687-691.

2. AGA Institute. AGA Institute medical position statement on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1977-1980.

3. Green PH, Jabri B. Coeliac disease. Lancet. 2003;362(9381):383-391.

4. Catassi C, Kryszak D, Louis-Jacques O, et al. Detection of celiac disease in primary care: a multicenter case-finding study in North America. Am J Gastroenterol. 2007;102(7);1454-1460.

5. Demirçeken FG, Kansu A, Kuloglu Z, et al. Human tissue transglutaminase antibody screening by immunochromatographic line immunoassay for early diagnosis of celiac disease in Turkish children. Turk J Gastroenterol. 2008;19(1):14-21.

6. van der Windt DA, Jellema P, Mulder CJ, et al. Diagnostic testing for celiac disease among patients with abdominal symptoms: a systematic review. JAMA. 2010;303(17):1738-1746.

7. Freeman HJ. Neurological disorders in adult celiac disease. Can J Gastroenterol. 2008; 22(11):909-911.

8. Tack GJ, Verbeek WHM, Schreurs MWJ, Mulder CJJ. The spectrum of celiac disease: epidemiology, clinical aspects and treatment. Nat Rev Gastroenterol Hepatol. 2010;7(4):204-213.

9. Farrell RJ, Kelly CP. Celiac sprue. N Engl J Med. 2002;346(3):180-188.

10. Green PHR, Cellier C. Celiac disease. N Engl J Med. 2007;357(17):1731-1743.

11. Fasano A, Berti I, Gerarduzzi T, et al. Prevalence of celiac disease in at-risk and not-at-risk groups in the United States: a large multicenter study. Arch Intern Med. 2003;163(3):286-292.

12. Sud S, Marcon M, Assor E, et al. Celiac disease and pediatric type 1 diabetes: diagnostic and treatment dilemmas. Int J Pediatr Endocrinol. 2010;2010:161285. Epub 2010 Jun 23.

13. Swigonski NL, Kuhlenschmidt HL, Bull MJ, et al. Screening for celiac disease in asymptomatic children with Down syndrome: cost-effectiveness of preventing lymphoma. Pediatrics. 2006;118(2):594-602.

14. Bonamico M, Pasquino AM, Mariani P, et al; Italian Society of Pediatric Gastroenterology Hepatology (SIGEP); Italian Study Group for Turner Syndrome (ISGTS). Prevalence and clinical picture of celiac disease in Turner syndrome. J Clin Endocrinol Metab. 2002;87(12): 5495-5498.

15. Giannotti A, Tiberio G, Castro M, et al. Coeliac disease in Williams syndrome. J Med Genet. 2001;38(11):767–768.

16. Alaedini A, Green P. Narrative review: celiac disease: understanding a complex autoimmune disorder. Ann Intern Med. 2005;142(4): 289-298.

17. Fröhlich-Reiterer EE, Hofer S, Kaspers S, et al. Screening frequency for celiac disease and autoimmune thyroiditis in children and adolescents with type 1 diabetes mellitus: data from a German/Austrian multicentre survey. Pediatr Diabetes. 2008;9(6):546-553.

18. American Diabetes Association. Standards of medical care in diabetes—2009. Diabetes Care. 2009;32(1):S13–S61.

19. Losowsky MS. A history of coeliac disease. Dig Dis. 2008;26(2):112-120.

20. Evans KE, Hadjivassilou M, Sanders DS. Understanding ‘silent’ coeliac disease: complications in diagnosis and treatment. Gastrointest Nurs. 2010;8(2):26-32.

21. Lurie Y, Landau DA, Pfeffer J, Oren R. Celiac disease diagnosed in the elderly. J Clin Gastroenterol. 2008;42(1):59-61.

22. Vilppula A, Kaukinen K, Luostarinen L, et al. Increasing prevalence and high incidence of celiac disease in elderly people: a population-based study. BMC Gastroenterol. 2009 Jun 29;9:49. 

23. Lo W, Sano K, Lebwohl B, et al. Changing presentation of adult celiac disease. Dig Dis Sci. 2003;48(2):395-398.

24. Alaedini A, Okamoto H, Briani C, et al. Immune cross-reactivity in celiac disease: anti-gliadin antibodies bind to neuronal synapsin I. J Immunol. 2007;178(10):6590-6595.

25. Tursi A, Giorgetti G, Brandimarte G, et al. Prevalence and clinical presentation of subclinical/silent celiac disease in adults: an analysis on a 12-year observation. Hepatogastroenterology. 2001;48(38):462-464.

26. Rostom A, Murray JA, Kagnoff MF. American Gastroenterological Association (AGA) Institute technical review on the diagnosis and management of celiac disease. Gastroenterology. 2006;131(6):1981-2002.

27. Ferguson A, Arranz E, O’Mahony S. Clinical and pathological spectrum of coeliac disease—active, silent, latent, potential. Gut. 1993;34(2): 150-151.

28. Logan RFA. Problems and pitfalls in epidemiological studies of coeliac disease. In: Auricchio S, Visakorpi JK, eds. Common Food Intolerances 1. Epidemiology of Coeliac Disease (Dynamic Nutrition Research)(Pt 1). Basel, Switzerland: Karger; 1992:14-22.

29. Marsh MN. Gluten, major histocompatibility complex, and the small intestine: a molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology. 1992;102(1):330-354.

30. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of coeliac disease: time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol. 1999;11(10):1185-1194.

31. Corazza GR, Villanaci V. Coeliac disease.
J Clin Pathol. 2005;58(6):573-574.

32. Hadithi M, von Blomberg BM, Crusius JB, et al. Accuracy of serologic tests and HLA-DQ typing for diagnosing celiac disease. Ann Intern Med. 2007;147(5):294-302.

33. Lerner A. New therapeutic strategies for celiac disease. Autoimmun Rev. 2010;9(3):144-147.

34. Catassi C, Fasano A. Celiac disease diagnosis: simple rules are better than complicated algorithms. Am J Med. 2010;123(8):691-693.

35. Rubio-Tapia A, Kyle RA, Kaplan EL, et al. Increased prevalence and mortality in undiagnosed celiac disease. Gastroenterology. 2009; 137(1):88–93.

36. Harewood GC, Holub JL, Lieberman DA. Variation in small bowel biopsy performance among diverse endoscopy settings: results from a national endoscopic database. Am J Gastroenterol. 2004;99(9):1790-1794.

37. Brar P, Lee S, Lewis S, et al. Budesonide in the treatment of refractory celiac disease. Am J Gastroenterol. 2007;102(10):2265-2269.

38. Al-Toma A, Verbeek WHM, Hadithi M, et al. Survival in refractory coeliac disease and enteropathy-associated T-cell lymphoma: retrospective evaluation of single-centre experience. Gut. 2007;56(10):1373-1378.

39. Kaukinen K, Mäki M, Partanen J, et al. Celiac disease without villous atrophy: revision of criteria called for. Dig Dis Sci. 2001;46(4):879-887.

40. Mohamed BM, Feighery C, Coates C, et al. The absence of a mucosal lesion on standard histological examination does not exclude diagnosis of celiac disease. Dig Dis Sci. 2008; 53(1):52-61.

41. Fasano A, Catassi C. Current approaches to diagnosis and treatment of celiac disease: an evolving spectrum. Gastroenterology. 2001; 120(3):636-651.

42. US Food and Drug Administration. Topic-specific labeling information (2010). www.fda.gov/Food/LabelingNutrition/FoodLabeling GuidanceRegulatoryInformation/Topic-Specific LabelingInformation/default.htm. Accessed March 28, 2011.

43. Lindfors K, Blomqvist T, Juuti-Uusitalo K, et al. Live probiotic Bifidobacterium lactis bacteria inhibit the toxic effects induced by wheat gliadin in epithelial cell culture. Clin Exp Immunol. 2008;152(3):552-558.

44. Cellier C, Delabesse E, Helmer C, et al; French Coeliac Disease Study Group. Refractory sprue, coeliac disease, and enteropathy-associated T-cell lymphoma. Lancet. 2000;356(9225): 203-208.

45. United European Gastroenterology. When is a coeliac a coeliac? Report of a working group of the United European Gastroenterology Week in Amsterdam, 2001. Eur J Gastroenterol Hepatol. 2001;13(9):1123-1128.

46. Malamut G, Afchain P, Verkarre V, et al. Presentation and long-term follow-up of refractory celiac disease: comparison of type I with type II. Gastroenterology. 2009;136(1):81-90.

47. Al-Toma A, Goerres MS, Meijer JW, et al. Human leukocyte antigen-DQ2 homozygosity and the development of refractory celiac disease and enteropathy-associated T-cell lymphoma. Clin Gastroenterol Hepatol. 2006;4(3): 315-319.

48. Mauriño E, Niveloni S, Cherñavsky A, et al. Azathioprine in refractory sprue: results from a prospective, open-label study. Am J Gastroenterol. 2002;97(10):2595–2602.

49. Tack GJ, Verbeek WHM, Al-Toma A, et al. Evaluation of cladribine treatment in refractory celiac disease type II. World J Gastroenterol. 2011;17(4):506–513.

50. Akobeng AK, Ramanan AV, Buchan I, Heller RF. Effect of breast feeding on risk of coeliac disease: a systematic review and meta-analysis of observational studies. Arch Dis Child. 2006; 91(1):39-43.

51. Rashid M, Butzner JD, Warren R, et al. Home blood testing for celiac disease: recommendations for management. Can Fam Physician. 2009;55(2):151-153.

52. Hershcovici T, Leshno M, Goldin E, et al. Cost effectiveness of mass screening for coeliac disease is determined by time-delay to diagnosis and quality of life on a gluten-free diet. Aliment Pharmacol Ther. 2010;31(8):901-910.

At age 82, Helen was healthy and active and living independently. A mother, grandmother, and great-grandmother, she enjoyed aerobics, tai chi, and walking, painting (which she also taught), writing poetry, and stimulating conversation. She took pride in looking much younger than her age and watched out for her older neighbors.

An active participant in her health care, Helen had been happy when, at 75, she was told by her primary care provider that she no longer needed regular mammograms. But one morning, seven years later, she felt a sharp pain in her right breast. Self-examination revealed a grape-sized lesion under her nipple. Helen sought immediate health care and was diagnosed with a stage IIb tumor.

Given an option of lumpectomy followed by radiation, Helen decided that a double mastectomy would better allow her to return to the life she had been living. After her surgery, however, Helen experienced a steady decline, with increasing pain, debilitating skin lesions, fractures, and edema. Against her will, she was moved to an assisted living facility, where she was too debilitated to participate in activities. Helen died six months later—three years after she discovered her breast lump.

The lack of clear breast cancer screening guidelines has left many providers confused about how to advise their patients, particularly women older than 75. Screening recommendations based on patient age alone are of insufficient value, as health status and life expectancy—which vary widely in this patient population—are also, along with patient preferences, important considerations. The purpose of this article is to present the reported benefits and risks of breast cancer screening among older women, in order to help primary care providers more effectively advise their elderly female patients in the decision-making process.

The Breast Cancer Screening Debate
There is strong consensus among expert advisory groups (the US Preventive Services Task Force,¹ the American Cancer Society,² the American College of Obstetricians and Gynecologists,³ the American Academy of Family Physicians,⁴ and the American Geriatrics Society⁵) that mammography is to be recommended to screen for the early detection of breast cancer in women between ages 50 and 75. However, a recent review of the randomized controlled studies on which these recommendations were based suggests that the benefits of mammography may be relatively minimal, and that the risks for overdiagnosis and overtreatment may be significant.⁶ None of these trials enrolled women older than 74, so further information is needed to make evidence-based decisions regarding breast cancer screening for the older woman. Currently available evidence for such decision making is limited to observational or retrospective analyses.⁶

Women 75 or older have a greater risk than younger women for breast cancer, but older women are also at greater risk for dying of another disease—even those who have breast cancer.^7,8 Thus, as with any health screening, it is advisable that a woman’s health status be carefully considered before screening decisions are made.

Breast Cancer in Older Women
Breast cancer incidence increases with age. Almost half of all invasive breast cancers and breast cancer deaths occur among women 65 and older, and almost one-quarter of all invasive breast cancers occur in women age 80 and older.⁹ Approximately one in six women diagnosed with breast cancer dies of the disease within 10 years.¹⁰ Once the cancer has metastasized, median survival time is two to four years. Older women have about a 1% chance of dying of breast cancer in a 10-year period.¹¹

At the same time, older women are more likely than their younger counterparts to have comorbid illnesses—conditions that can compete with breast cancer as a cause of death and increase treatment-related morbidity. Older women with breast cancer and two or more other comorbid diseases are 20 times more likely to die of one of these diseases than of breast cancer.⁸

Use of Mammography in Older Women
Regular mammography has been shown to help detect tumors that are smaller and at an earlier stage of development than would be identified without this screening.⁸ Women who undergo regular mammograms have been shown to outlive those who do not, provided they live for at least five years after starting screening.¹²

Life expectancy for older women is often underestimated. The average 75-year-old woman has 12 more years of life expectancy, and the healthiest 25% of 80-year-old women will live an additional 13 years.¹³ Most of these women do not receive regular mammographic screening for breast cancer. However, if health status and life expectancy were considered as screening criteria in addition to age, many of these women would qualify for mammographic screening.

Using the Surveillance, Epidemiology, and End Results (SEER) Medicare database to evaluate 12,358 women age 80 and older who had been diagnosed with breast cancer, Badgwell et al¹⁴ found that only 51% had had one or more mammograms within the previous five years. Biennial screening rates were 24% to 27% among women ages 80 to 84 and 14% to 23% among those 85 to 89. In this cohort of elderly women, the researchers found that regular mammographic screening was associated with detection of breast cancer at earlier stages and suggest that the low rates of screening in this age-group may represent a missed opportunity for early detection of disease.

In the US, 57% of breast cancers are detected by mammography; in the remaining cases, patients present with a palpable mass or in response to other breast symptoms. Cancers detected by mammography are found at an earlier stage.¹⁵

When statisticians used several models to assess the role of screening mammography in reducing breast cancer mortality, they found that mammography has contributed to about 46% of the overall reduction in breast cancer deaths in the US during the past 20 years.¹⁶ Mammography is more effective in detecting breast cancers in older than younger women, and the number of false-positive results decreases among women of advancing age. The sensitivity and specificity of mammography for detection of breast cancer is 85% and 94%, respectively, for women ages 75 to 89.¹⁷

Although this strong specificity would seem to suggest that older women are at low risk for overdiagnosis, this is not the case. Many benign and clinically insignificant lesions are also detected through mammography, resulting in unnecessary breast biopsies. In one group of 23,000 women age 65 or older who underwent a one-time screening mammogram, 8% had an abnormal result that required additional evaluation. Among these women, only about 10% actually had cancer.¹⁸ Rates of false-positive mammographic findings vary by radiologist but are generally higher among women who are younger than 65.¹⁹

Schonberg²⁰ followed 2,011 community-dwelling women 80 and older who underwent mammography screening and found an 11% rate of false-positive results. Ductal carcinoma in situ (DCIS) is a common mammographic finding in older women. Since only one-third of these cases will convert to invasive breast cancer over 10 to 15 years, DCIS likely exemplifies overdiagnosis in older women.²⁰

Clinical Breast Exams and Self-Examination
There is insufficient evidence to determine whether clinical breast exams improve early detection and treatment outcomes in women with breast cancer. Additionally, results from two large randomized controlled trials of breast self-examination suggest that the practice is not of benefit in reducing breast cancer mortality and morbidity.^1,21,22

Age, Breast Cancer Types, and Outcomes
It has been suggested that older women may be subject to less aggressive cancers and thus be more vulnerable to overdiagnosis.²³ Schonberg et al²⁴ evaluated SEER data to determine the tumor characteristics, treatments, and outcomes in women 80 and older, compared with women ages 67 to 79. They found no difference in tumor grade or hormone receptivity between these groups. It is important to note that women older than 80 were significantly more likely than younger women to die of breast cancer, perhaps in part because the older patients were less likely to receive aggressive treatment (see “Breast Cancer Treatment in Older Women”^24-26).

Women between ages 74 and 85 who undergo regular mammographic screenings have been shown to have half the risk for breast cancer–related death, compared with those who are not screened.²⁵ However, risks have been shown to outweigh benefits when mammography is continued into old age without regard to life expectancy.

Walter et al²⁷ studied a group of 216 frail, nursing home–­eligible older women who had had at least one mammogram. Seventeen percent had abnormal results, and most opted for further evaluation with breast biopsy. Of these biopsies, 23% yielded positive results, and of these, 75% revealed invasive breast cancer; the remaining 25% of women had DCIS.

All of the women with abnormal biopsy results underwent surgical treatment, but half died of other causes or experienced surgical complications. The investigators found that 1% of the women may have received some benefit from screening, but more women experienced harm as a result of the mammogram and subsequent procedures. The study authors, along with almost all clinicians who have written on this topic, agree that mammography is inappropriate for frail, elderly women with less than five years’ life expectancy.²⁷

Expert Guideline Recommendations
Provider recommendation has been found to be the most important factor in older women’s decisions to have or forego a mammogram.⁸ Unfortunately, there is little clear, decisive support for providers to help women make this decision. The summary of breast cancer screening recommendations for older women shown in the table reveals that no consensus exists among the expert panels regarding the best approach. In part, these discrepancies can be explained by the low numbers of women older than 75 who have been included in clinical trials evaluating the risks and benefits of mammography screening.

Helpful Criteria for Breast Cancer Screening Decisions
In 2003, the American Geriatrics Society Ethics Committee published a set of basic rules to guide decision making regarding screening tests.⁵ These include the following:

• In patients with a limited life expectancy, focus should be on treatments that are likely to offer immediate benefit

• Patients with dementia or multiple comorbidities may find routine screening tests burdensome

• Screening decisions should be individualized rather than based on age alone

• Health care systems and insurance plans should not restrict coverage for screening tests in older adults based solely on age.⁵

Walter and Covinsky¹³ created a model to facilitate decisions about mammography for older women. They suggest first estimating life expectancy according to age and health status (see Arias²⁸ at www.cdc.gov/nchs/data/nvsr/nvsr58/nvsr58_21.pdf). Factors considered in this decision model include a woman’s risk for dying of breast cancer, the effect of screening on this death rate, the potential risks involved with screening, and the woman’s preferences.

According to this model, Helen, at 82, would have had a life expectancy of 11.5 years. The annual breast cancer mortality rate for a woman of her age is 157/100,000.²⁹ Based on these data, Helen’s risk for dying of breast cancer is calculated to be 1.8%. The number of patients Helen’s age who would have to be screened to prevent one case of breast cancer is 240.

In Helen’s case, the risks involved with screening mammography would include a roughly 8% chance of her needing a subsequent diagnostic mammogram and/or breast biopsy. If Helen underwent biopsy, there would be a 75% chance that the suspicious mammogram would prove to be a false-negative result, possibly causing the patient undue anxiety. In Walter and Covinski’s model,¹³ these possibilities would be discussed with Helen in advance to help her clarify her own values and reach a screening decision.

By comparison, consider a hypothetical 70-year-old woman who is in the lowest quadrant of health status for her age; based on her age alone, the expert panels would agree that screening mammography is indicated. This patient has a life expectancy of 9.5 years and a 1.2% lifetime risk for dying of breast cancer. To prevent one case of breast cancer, 642 women of this age and health status would have to be screened. Thus, Helen would derive far more benefit from screening than would this hypothetical woman.

Life Expectancy, Health Status
In order to help an older woman make well-reasoned decisions about mammography screening, it is important for the clinician to evaluate her overall health status and life expectancy, as well as her risk for breast cancer and her long-term goals.

The potential benefit of any preventive health screening can be reduced in patients with a limited life span. Schonberg,²⁰ who created a nine-item tool to predict four-year mortality, demonstrated that women with a score higher than the cut-point of 14 were unlikely to benefit from mammography screening, based on their limited life expectancy.

According to Schonberg,²⁰ mortality risk increases with advancing age, male gender, and low BMI; a diagnosis of diabetes, cancer, or COPD; smoking, functional limitations, poor self-rated health, and a recent history of hospitalization. In Helen’s case, a BMI of 25, her nonsmoking status, an absence of significant comorbidities, and good functional status result in a score of 5 on the Schonberg tool. The associated life expectancy of greater than four years would have made mammography screening an advisable option for her.

Several easily administered tools are available to predict mortality or vulnerability. The Vulnerable Elders Survey (VES-13),^30,31 for example, is a 13-item tool developed by researchers from the RAND Health Project, the University of California–Los Angeles, and the Veterans Administration to predict vulnerability among older adults. Using a cut-point of 3, researchers who conducted this survey identified one-third of a cohort of 6,205 community-dwelling elders as vulnerable—that is, at four times the risk for functional decline or death, compared with the rest of the elderly population.³¹ Helen, with a VES-13 score of 0, would not have been considered vulnerable.

Additionally, risk for breast cancer can be calculated in the office setting with the National Cancer Institute’s Breast Cancer Risk Calculator³² (see www.cancer.gov/bcrisktool). Breast cancer risk factors for the older woman that are not included in the calculator but that may further influence decision making include:

• Use of hormone replacement therapy

• Obesity

• Increased bone mineral density.³³

It is important to determine a woman’s willingness to undergo mammography and any indicated follow-up procedures. Screening decisions can also be facilitated by discussions about the patient’s values regarding increased longevity, quality-of-life goals, pain and symptom management, and her available support systems.

Conclusion
Decisions about breast cancer screening for older women should be influenced by the health status and life expectancy of each patient, her goals for her remaining years, and her risk for breast cancer. Treatment decisions, when necessary, must take these factors, as well as severity of disease, into account. The reduction in quality of life that is inevitable with a diagnosis of advanced-stage metastatic breast cancer should be factored into the decision.

Helen, whose story began this article, was a real patient. Given her previous good health, she hardly expected her final years to be defined by the pain and disability of metastatic breast cancer. Her case illustrates the potential consequences of restricting breast cancer screening decisions to age-based recommendations alone. 

At age 82, Helen was healthy and active and living independently. A mother, grandmother, and great-grandmother, she enjoyed aerobics, tai chi, and walking, painting (which she also taught), writing poetry, and stimulating conversation. She took pride in looking much younger than her age and watched out for her older neighbors.

An active participant in her health care, Helen had been happy when, at 75, she was told by her primary care provider that she no longer needed regular mammograms. But one morning, seven years later, she felt a sharp pain in her right breast. Self-examination revealed a grape-sized lesion under her nipple. Helen sought immediate health care and was diagnosed with a stage IIb tumor.

Given an option of lumpectomy followed by radiation, Helen decided that a double mastectomy would better allow her to return to the life she had been living. After her surgery, however, Helen experienced a steady decline, with increasing pain, debilitating skin lesions, fractures, and edema. Against her will, she was moved to an assisted living facility, where she was too debilitated to participate in activities. Helen died six months later—three years after she discovered her breast lump.

The lack of clear breast cancer screening guidelines has left many providers confused about how to advise their patients, particularly women older than 75. Screening recommendations based on patient age alone are of insufficient value, as health status and life expectancy—which vary widely in this patient population—are also, along with patient preferences, important considerations. The purpose of this article is to present the reported benefits and risks of breast cancer screening among older women, in order to help primary care providers more effectively advise their elderly female patients in the decision-making process.

The Breast Cancer Screening Debate
There is strong consensus among expert advisory groups (the US Preventive Services Task Force,¹ the American Cancer Society,² the American College of Obstetricians and Gynecologists,³ the American Academy of Family Physicians,⁴ and the American Geriatrics Society⁵) that mammography is to be recommended to screen for the early detection of breast cancer in women between ages 50 and 75. However, a recent review of the randomized controlled studies on which these recommendations were based suggests that the benefits of mammography may be relatively minimal, and that the risks for overdiagnosis and overtreatment may be significant.⁶ None of these trials enrolled women older than 74, so further information is needed to make evidence-based decisions regarding breast cancer screening for the older woman. Currently available evidence for such decision making is limited to observational or retrospective analyses.⁶

Women 75 or older have a greater risk than younger women for breast cancer, but older women are also at greater risk for dying of another disease—even those who have breast cancer.^7,8 Thus, as with any health screening, it is advisable that a woman’s health status be carefully considered before screening decisions are made.

Breast Cancer in Older Women
Breast cancer incidence increases with age. Almost half of all invasive breast cancers and breast cancer deaths occur among women 65 and older, and almost one-quarter of all invasive breast cancers occur in women age 80 and older.⁹ Approximately one in six women diagnosed with breast cancer dies of the disease within 10 years.¹⁰ Once the cancer has metastasized, median survival time is two to four years. Older women have about a 1% chance of dying of breast cancer in a 10-year period.¹¹

At the same time, older women are more likely than their younger counterparts to have comorbid illnesses—conditions that can compete with breast cancer as a cause of death and increase treatment-related morbidity. Older women with breast cancer and two or more other comorbid diseases are 20 times more likely to die of one of these diseases than of breast cancer.⁸

Use of Mammography in Older Women
Regular mammography has been shown to help detect tumors that are smaller and at an earlier stage of development than would be identified without this screening.⁸ Women who undergo regular mammograms have been shown to outlive those who do not, provided they live for at least five years after starting screening.¹²

Life expectancy for older women is often underestimated. The average 75-year-old woman has 12 more years of life expectancy, and the healthiest 25% of 80-year-old women will live an additional 13 years.¹³ Most of these women do not receive regular mammographic screening for breast cancer. However, if health status and life expectancy were considered as screening criteria in addition to age, many of these women would qualify for mammographic screening.

Using the Surveillance, Epidemiology, and End Results (SEER) Medicare database to evaluate 12,358 women age 80 and older who had been diagnosed with breast cancer, Badgwell et al¹⁴ found that only 51% had had one or more mammograms within the previous five years. Biennial screening rates were 24% to 27% among women ages 80 to 84 and 14% to 23% among those 85 to 89. In this cohort of elderly women, the researchers found that regular mammographic screening was associated with detection of breast cancer at earlier stages and suggest that the low rates of screening in this age-group may represent a missed opportunity for early detection of disease.

In the US, 57% of breast cancers are detected by mammography; in the remaining cases, patients present with a palpable mass or in response to other breast symptoms. Cancers detected by mammography are found at an earlier stage.¹⁵

When statisticians used several models to assess the role of screening mammography in reducing breast cancer mortality, they found that mammography has contributed to about 46% of the overall reduction in breast cancer deaths in the US during the past 20 years.¹⁶ Mammography is more effective in detecting breast cancers in older than younger women, and the number of false-positive results decreases among women of advancing age. The sensitivity and specificity of mammography for detection of breast cancer is 85% and 94%, respectively, for women ages 75 to 89.¹⁷

Although this strong specificity would seem to suggest that older women are at low risk for overdiagnosis, this is not the case. Many benign and clinically insignificant lesions are also detected through mammography, resulting in unnecessary breast biopsies. In one group of 23,000 women age 65 or older who underwent a one-time screening mammogram, 8% had an abnormal result that required additional evaluation. Among these women, only about 10% actually had cancer.¹⁸ Rates of false-positive mammographic findings vary by radiologist but are generally higher among women who are younger than 65.¹⁹

Schonberg²⁰ followed 2,011 community-dwelling women 80 and older who underwent mammography screening and found an 11% rate of false-positive results. Ductal carcinoma in situ (DCIS) is a common mammographic finding in older women. Since only one-third of these cases will convert to invasive breast cancer over 10 to 15 years, DCIS likely exemplifies overdiagnosis in older women.²⁰

Clinical Breast Exams and Self-Examination
There is insufficient evidence to determine whether clinical breast exams improve early detection and treatment outcomes in women with breast cancer. Additionally, results from two large randomized controlled trials of breast self-examination suggest that the practice is not of benefit in reducing breast cancer mortality and morbidity.^1,21,22

Age, Breast Cancer Types, and Outcomes
It has been suggested that older women may be subject to less aggressive cancers and thus be more vulnerable to overdiagnosis.²³ Schonberg et al²⁴ evaluated SEER data to determine the tumor characteristics, treatments, and outcomes in women 80 and older, compared with women ages 67 to 79. They found no difference in tumor grade or hormone receptivity between these groups. It is important to note that women older than 80 were significantly more likely than younger women to die of breast cancer, perhaps in part because the older patients were less likely to receive aggressive treatment (see “Breast Cancer Treatment in Older Women”^24-26).

Women between ages 74 and 85 who undergo regular mammographic screenings have been shown to have half the risk for breast cancer–related death, compared with those who are not screened.²⁵ However, risks have been shown to outweigh benefits when mammography is continued into old age without regard to life expectancy.

Walter et al²⁷ studied a group of 216 frail, nursing home–­eligible older women who had had at least one mammogram. Seventeen percent had abnormal results, and most opted for further evaluation with breast biopsy. Of these biopsies, 23% yielded positive results, and of these, 75% revealed invasive breast cancer; the remaining 25% of women had DCIS.

All of the women with abnormal biopsy results underwent surgical treatment, but half died of other causes or experienced surgical complications. The investigators found that 1% of the women may have received some benefit from screening, but more women experienced harm as a result of the mammogram and subsequent procedures. The study authors, along with almost all clinicians who have written on this topic, agree that mammography is inappropriate for frail, elderly women with less than five years’ life expectancy.²⁷

Expert Guideline Recommendations
Provider recommendation has been found to be the most important factor in older women’s decisions to have or forego a mammogram.⁸ Unfortunately, there is little clear, decisive support for providers to help women make this decision. The summary of breast cancer screening recommendations for older women shown in the table reveals that no consensus exists among the expert panels regarding the best approach. In part, these discrepancies can be explained by the low numbers of women older than 75 who have been included in clinical trials evaluating the risks and benefits of mammography screening.

Helpful Criteria for Breast Cancer Screening Decisions
In 2003, the American Geriatrics Society Ethics Committee published a set of basic rules to guide decision making regarding screening tests.⁵ These include the following:

• In patients with a limited life expectancy, focus should be on treatments that are likely to offer immediate benefit

• Patients with dementia or multiple comorbidities may find routine screening tests burdensome

• Screening decisions should be individualized rather than based on age alone

• Health care systems and insurance plans should not restrict coverage for screening tests in older adults based solely on age.⁵

Walter and Covinsky¹³ created a model to facilitate decisions about mammography for older women. They suggest first estimating life expectancy according to age and health status (see Arias²⁸ at www.cdc.gov/nchs/data/nvsr/nvsr58/nvsr58_21.pdf). Factors considered in this decision model include a woman’s risk for dying of breast cancer, the effect of screening on this death rate, the potential risks involved with screening, and the woman’s preferences.

According to this model, Helen, at 82, would have had a life expectancy of 11.5 years. The annual breast cancer mortality rate for a woman of her age is 157/100,000.²⁹ Based on these data, Helen’s risk for dying of breast cancer is calculated to be 1.8%. The number of patients Helen’s age who would have to be screened to prevent one case of breast cancer is 240.

In Helen’s case, the risks involved with screening mammography would include a roughly 8% chance of her needing a subsequent diagnostic mammogram and/or breast biopsy. If Helen underwent biopsy, there would be a 75% chance that the suspicious mammogram would prove to be a false-negative result, possibly causing the patient undue anxiety. In Walter and Covinski’s model,¹³ these possibilities would be discussed with Helen in advance to help her clarify her own values and reach a screening decision.

By comparison, consider a hypothetical 70-year-old woman who is in the lowest quadrant of health status for her age; based on her age alone, the expert panels would agree that screening mammography is indicated. This patient has a life expectancy of 9.5 years and a 1.2% lifetime risk for dying of breast cancer. To prevent one case of breast cancer, 642 women of this age and health status would have to be screened. Thus, Helen would derive far more benefit from screening than would this hypothetical woman.

Life Expectancy, Health Status
In order to help an older woman make well-reasoned decisions about mammography screening, it is important for the clinician to evaluate her overall health status and life expectancy, as well as her risk for breast cancer and her long-term goals.

The potential benefit of any preventive health screening can be reduced in patients with a limited life span. Schonberg,²⁰ who created a nine-item tool to predict four-year mortality, demonstrated that women with a score higher than the cut-point of 14 were unlikely to benefit from mammography screening, based on their limited life expectancy.

According to Schonberg,²⁰ mortality risk increases with advancing age, male gender, and low BMI; a diagnosis of diabetes, cancer, or COPD; smoking, functional limitations, poor self-rated health, and a recent history of hospitalization. In Helen’s case, a BMI of 25, her nonsmoking status, an absence of significant comorbidities, and good functional status result in a score of 5 on the Schonberg tool. The associated life expectancy of greater than four years would have made mammography screening an advisable option for her.

Several easily administered tools are available to predict mortality or vulnerability. The Vulnerable Elders Survey (VES-13),^30,31 for example, is a 13-item tool developed by researchers from the RAND Health Project, the University of California–Los Angeles, and the Veterans Administration to predict vulnerability among older adults. Using a cut-point of 3, researchers who conducted this survey identified one-third of a cohort of 6,205 community-dwelling elders as vulnerable—that is, at four times the risk for functional decline or death, compared with the rest of the elderly population.³¹ Helen, with a VES-13 score of 0, would not have been considered vulnerable.

Additionally, risk for breast cancer can be calculated in the office setting with the National Cancer Institute’s Breast Cancer Risk Calculator³² (see www.cancer.gov/bcrisktool). Breast cancer risk factors for the older woman that are not included in the calculator but that may further influence decision making include:

• Use of hormone replacement therapy

• Obesity

• Increased bone mineral density.³³

It is important to determine a woman’s willingness to undergo mammography and any indicated follow-up procedures. Screening decisions can also be facilitated by discussions about the patient’s values regarding increased longevity, quality-of-life goals, pain and symptom management, and her available support systems.

Conclusion
Decisions about breast cancer screening for older women should be influenced by the health status and life expectancy of each patient, her goals for her remaining years, and her risk for breast cancer. Treatment decisions, when necessary, must take these factors, as well as severity of disease, into account. The reduction in quality of life that is inevitable with a diagnosis of advanced-stage metastatic breast cancer should be factored into the decision.

Helen, whose story began this article, was a real patient. Given her previous good health, she hardly expected her final years to be defined by the pain and disability of metastatic breast cancer. Her case illustrates the potential consequences of restricting breast cancer screening decisions to age-based recommendations alone. 

At age 82, Helen was healthy and active and living independently. A mother, grandmother, and great-grandmother, she enjoyed aerobics, tai chi, and walking, painting (which she also taught), writing poetry, and stimulating conversation. She took pride in looking much younger than her age and watched out for her older neighbors.

An active participant in her health care, Helen had been happy when, at 75, she was told by her primary care provider that she no longer needed regular mammograms. But one morning, seven years later, she felt a sharp pain in her right breast. Self-examination revealed a grape-sized lesion under her nipple. Helen sought immediate health care and was diagnosed with a stage IIb tumor.

Given an option of lumpectomy followed by radiation, Helen decided that a double mastectomy would better allow her to return to the life she had been living. After her surgery, however, Helen experienced a steady decline, with increasing pain, debilitating skin lesions, fractures, and edema. Against her will, she was moved to an assisted living facility, where she was too debilitated to participate in activities. Helen died six months later—three years after she discovered her breast lump.

The lack of clear breast cancer screening guidelines has left many providers confused about how to advise their patients, particularly women older than 75. Screening recommendations based on patient age alone are of insufficient value, as health status and life expectancy—which vary widely in this patient population—are also, along with patient preferences, important considerations. The purpose of this article is to present the reported benefits and risks of breast cancer screening among older women, in order to help primary care providers more effectively advise their elderly female patients in the decision-making process.

The Breast Cancer Screening Debate
There is strong consensus among expert advisory groups (the US Preventive Services Task Force,¹ the American Cancer Society,² the American College of Obstetricians and Gynecologists,³ the American Academy of Family Physicians,⁴ and the American Geriatrics Society⁵) that mammography is to be recommended to screen for the early detection of breast cancer in women between ages 50 and 75. However, a recent review of the randomized controlled studies on which these recommendations were based suggests that the benefits of mammography may be relatively minimal, and that the risks for overdiagnosis and overtreatment may be significant.⁶ None of these trials enrolled women older than 74, so further information is needed to make evidence-based decisions regarding breast cancer screening for the older woman. Currently available evidence for such decision making is limited to observational or retrospective analyses.⁶

Women 75 or older have a greater risk than younger women for breast cancer, but older women are also at greater risk for dying of another disease—even those who have breast cancer.^7,8 Thus, as with any health screening, it is advisable that a woman’s health status be carefully considered before screening decisions are made.

Breast Cancer in Older Women
Breast cancer incidence increases with age. Almost half of all invasive breast cancers and breast cancer deaths occur among women 65 and older, and almost one-quarter of all invasive breast cancers occur in women age 80 and older.⁹ Approximately one in six women diagnosed with breast cancer dies of the disease within 10 years.¹⁰ Once the cancer has metastasized, median survival time is two to four years. Older women have about a 1% chance of dying of breast cancer in a 10-year period.¹¹

At the same time, older women are more likely than their younger counterparts to have comorbid illnesses—conditions that can compete with breast cancer as a cause of death and increase treatment-related morbidity. Older women with breast cancer and two or more other comorbid diseases are 20 times more likely to die of one of these diseases than of breast cancer.⁸

Use of Mammography in Older Women
Regular mammography has been shown to help detect tumors that are smaller and at an earlier stage of development than would be identified without this screening.⁸ Women who undergo regular mammograms have been shown to outlive those who do not, provided they live for at least five years after starting screening.¹²

Life expectancy for older women is often underestimated. The average 75-year-old woman has 12 more years of life expectancy, and the healthiest 25% of 80-year-old women will live an additional 13 years.¹³ Most of these women do not receive regular mammographic screening for breast cancer. However, if health status and life expectancy were considered as screening criteria in addition to age, many of these women would qualify for mammographic screening.

Using the Surveillance, Epidemiology, and End Results (SEER) Medicare database to evaluate 12,358 women age 80 and older who had been diagnosed with breast cancer, Badgwell et al¹⁴ found that only 51% had had one or more mammograms within the previous five years. Biennial screening rates were 24% to 27% among women ages 80 to 84 and 14% to 23% among those 85 to 89. In this cohort of elderly women, the researchers found that regular mammographic screening was associated with detection of breast cancer at earlier stages and suggest that the low rates of screening in this age-group may represent a missed opportunity for early detection of disease.

In the US, 57% of breast cancers are detected by mammography; in the remaining cases, patients present with a palpable mass or in response to other breast symptoms. Cancers detected by mammography are found at an earlier stage.¹⁵

When statisticians used several models to assess the role of screening mammography in reducing breast cancer mortality, they found that mammography has contributed to about 46% of the overall reduction in breast cancer deaths in the US during the past 20 years.¹⁶ Mammography is more effective in detecting breast cancers in older than younger women, and the number of false-positive results decreases among women of advancing age. The sensitivity and specificity of mammography for detection of breast cancer is 85% and 94%, respectively, for women ages 75 to 89.¹⁷

Although this strong specificity would seem to suggest that older women are at low risk for overdiagnosis, this is not the case. Many benign and clinically insignificant lesions are also detected through mammography, resulting in unnecessary breast biopsies. In one group of 23,000 women age 65 or older who underwent a one-time screening mammogram, 8% had an abnormal result that required additional evaluation. Among these women, only about 10% actually had cancer.¹⁸ Rates of false-positive mammographic findings vary by radiologist but are generally higher among women who are younger than 65.¹⁹

Schonberg²⁰ followed 2,011 community-dwelling women 80 and older who underwent mammography screening and found an 11% rate of false-positive results. Ductal carcinoma in situ (DCIS) is a common mammographic finding in older women. Since only one-third of these cases will convert to invasive breast cancer over 10 to 15 years, DCIS likely exemplifies overdiagnosis in older women.²⁰

Clinical Breast Exams and Self-Examination
There is insufficient evidence to determine whether clinical breast exams improve early detection and treatment outcomes in women with breast cancer. Additionally, results from two large randomized controlled trials of breast self-examination suggest that the practice is not of benefit in reducing breast cancer mortality and morbidity.^1,21,22

Age, Breast Cancer Types, and Outcomes
It has been suggested that older women may be subject to less aggressive cancers and thus be more vulnerable to overdiagnosis.²³ Schonberg et al²⁴ evaluated SEER data to determine the tumor characteristics, treatments, and outcomes in women 80 and older, compared with women ages 67 to 79. They found no difference in tumor grade or hormone receptivity between these groups. It is important to note that women older than 80 were significantly more likely than younger women to die of breast cancer, perhaps in part because the older patients were less likely to receive aggressive treatment (see “Breast Cancer Treatment in Older Women”^24-26).

Women between ages 74 and 85 who undergo regular mammographic screenings have been shown to have half the risk for breast cancer–related death, compared with those who are not screened.²⁵ However, risks have been shown to outweigh benefits when mammography is continued into old age without regard to life expectancy.

Walter et al²⁷ studied a group of 216 frail, nursing home–­eligible older women who had had at least one mammogram. Seventeen percent had abnormal results, and most opted for further evaluation with breast biopsy. Of these biopsies, 23% yielded positive results, and of these, 75% revealed invasive breast cancer; the remaining 25% of women had DCIS.

All of the women with abnormal biopsy results underwent surgical treatment, but half died of other causes or experienced surgical complications. The investigators found that 1% of the women may have received some benefit from screening, but more women experienced harm as a result of the mammogram and subsequent procedures. The study authors, along with almost all clinicians who have written on this topic, agree that mammography is inappropriate for frail, elderly women with less than five years’ life expectancy.²⁷

Expert Guideline Recommendations
Provider recommendation has been found to be the most important factor in older women’s decisions to have or forego a mammogram.⁸ Unfortunately, there is little clear, decisive support for providers to help women make this decision. The summary of breast cancer screening recommendations for older women shown in the table reveals that no consensus exists among the expert panels regarding the best approach. In part, these discrepancies can be explained by the low numbers of women older than 75 who have been included in clinical trials evaluating the risks and benefits of mammography screening.

Helpful Criteria for Breast Cancer Screening Decisions
In 2003, the American Geriatrics Society Ethics Committee published a set of basic rules to guide decision making regarding screening tests.⁵ These include the following:

• In patients with a limited life expectancy, focus should be on treatments that are likely to offer immediate benefit

• Patients with dementia or multiple comorbidities may find routine screening tests burdensome

• Screening decisions should be individualized rather than based on age alone

• Health care systems and insurance plans should not restrict coverage for screening tests in older adults based solely on age.⁵

Walter and Covinsky¹³ created a model to facilitate decisions about mammography for older women. They suggest first estimating life expectancy according to age and health status (see Arias²⁸ at www.cdc.gov/nchs/data/nvsr/nvsr58/nvsr58_21.pdf). Factors considered in this decision model include a woman’s risk for dying of breast cancer, the effect of screening on this death rate, the potential risks involved with screening, and the woman’s preferences.

According to this model, Helen, at 82, would have had a life expectancy of 11.5 years. The annual breast cancer mortality rate for a woman of her age is 157/100,000.²⁹ Based on these data, Helen’s risk for dying of breast cancer is calculated to be 1.8%. The number of patients Helen’s age who would have to be screened to prevent one case of breast cancer is 240.

In Helen’s case, the risks involved with screening mammography would include a roughly 8% chance of her needing a subsequent diagnostic mammogram and/or breast biopsy. If Helen underwent biopsy, there would be a 75% chance that the suspicious mammogram would prove to be a false-negative result, possibly causing the patient undue anxiety. In Walter and Covinski’s model,¹³ these possibilities would be discussed with Helen in advance to help her clarify her own values and reach a screening decision.

By comparison, consider a hypothetical 70-year-old woman who is in the lowest quadrant of health status for her age; based on her age alone, the expert panels would agree that screening mammography is indicated. This patient has a life expectancy of 9.5 years and a 1.2% lifetime risk for dying of breast cancer. To prevent one case of breast cancer, 642 women of this age and health status would have to be screened. Thus, Helen would derive far more benefit from screening than would this hypothetical woman.

Life Expectancy, Health Status
In order to help an older woman make well-reasoned decisions about mammography screening, it is important for the clinician to evaluate her overall health status and life expectancy, as well as her risk for breast cancer and her long-term goals.

The potential benefit of any preventive health screening can be reduced in patients with a limited life span. Schonberg,²⁰ who created a nine-item tool to predict four-year mortality, demonstrated that women with a score higher than the cut-point of 14 were unlikely to benefit from mammography screening, based on their limited life expectancy.

According to Schonberg,²⁰ mortality risk increases with advancing age, male gender, and low BMI; a diagnosis of diabetes, cancer, or COPD; smoking, functional limitations, poor self-rated health, and a recent history of hospitalization. In Helen’s case, a BMI of 25, her nonsmoking status, an absence of significant comorbidities, and good functional status result in a score of 5 on the Schonberg tool. The associated life expectancy of greater than four years would have made mammography screening an advisable option for her.

Several easily administered tools are available to predict mortality or vulnerability. The Vulnerable Elders Survey (VES-13),^30,31 for example, is a 13-item tool developed by researchers from the RAND Health Project, the University of California–Los Angeles, and the Veterans Administration to predict vulnerability among older adults. Using a cut-point of 3, researchers who conducted this survey identified one-third of a cohort of 6,205 community-dwelling elders as vulnerable—that is, at four times the risk for functional decline or death, compared with the rest of the elderly population.³¹ Helen, with a VES-13 score of 0, would not have been considered vulnerable.

Additionally, risk for breast cancer can be calculated in the office setting with the National Cancer Institute’s Breast Cancer Risk Calculator³² (see www.cancer.gov/bcrisktool). Breast cancer risk factors for the older woman that are not included in the calculator but that may further influence decision making include:

• Use of hormone replacement therapy

• Obesity

• Increased bone mineral density.³³

It is important to determine a woman’s willingness to undergo mammography and any indicated follow-up procedures. Screening decisions can also be facilitated by discussions about the patient’s values regarding increased longevity, quality-of-life goals, pain and symptom management, and her available support systems.

Conclusion
Decisions about breast cancer screening for older women should be influenced by the health status and life expectancy of each patient, her goals for her remaining years, and her risk for breast cancer. Treatment decisions, when necessary, must take these factors, as well as severity of disease, into account. The reduction in quality of life that is inevitable with a diagnosis of advanced-stage metastatic breast cancer should be factored into the decision.

Helen, whose story began this article, was a real patient. Given her previous good health, she hardly expected her final years to be defined by the pain and disability of metastatic breast cancer. Her case illustrates the potential consequences of restricting breast cancer screening decisions to age-based recommendations alone.