Clinical Review

Ob/Gyns are well trained to care for more than 99% of women with osteopenia and osteoporosis. Yet most women at risk are not being diagnosed or treated. Since recent publication of randomized clinical trials, however, we have clearer direction on the decisive diagnosis and management questions.

Who should have a bone mass test?

Ob/Gyns know best

Solomon DH, Connelly MT, Rosen CJ, et al. Factors related to the use of bone densitometry: survey responses of 494 primary care physicians in New England. Osteoporosis Int. 2003;14:123–129.

Ob/Gyns were significantly more likely to order bone densitometry than internists and family physicians. However, any physician, including Ob/Gyns, who believed (mistakenly) that calcium-plus-vitamin D effectively treats osteoporosis or that osteoporosis should not be diagnosed by densitometry used screening less frequently than physicians without those beliefs.

Any woman should have bone density testing if it might influence her medical care. Osteoporosis is notoriously difficult to diagnose without densitometry.¹

Besides identifying women at risk of fracture due to osteopenia or osteoporosis who are good candidates for drug therapy, testing can help motivate women to stop smoking, exercise, and take calcium and vitamin D to prevent bone loss. Just as women should know their weight, serum cholesterol, blood pressure, and mammographic findings, hypoestrogenic women should know their T-score.

The issue of relative costs and benefits of bone density testing is complex and continues to evolve. Although national organizations have guidelines (TABLE 1), it is not clear if their sensitivity and specificity are optimal for identifying appropriate candidates for screening.

A reliable, quick tool identifies who to test

Cadarette SM, Jaglal SB, Murray TM, McIsaac WJ, Lawrence L, Brown JP. Evaluation of decision rules for referring women for bone density by dual-energy X-ray absorptiometry. JAMA. 2001;286:57–63.

A 3-item Osteoporosis Risk Assessment Instrument was more sensitive and specific in identifying screening candidates than the National Osteoporosis Foundation (NOF) criteria, according to an analysis of screening algorithms used in 2,365 menopausal women in the Canadian Multicentre Osteoporosis Study. A simple calculation based on age, weight, and estrogen use (TABLE 2) was clinically applicable.

In my practice, I focus on all women who have been hypoestrogenic for 12 to 24 months regardless of age, women with a previous low-trauma fracture, and women who weigh less than 132 pounds. Evidence is mounting that early treatment of bone loss is the best way to prevent future fracture. Well before osteoporosis is detected, significant structural integrity of the spine and hip has been lost.

It is my belief that guidelines will ultimately recommend bone mineral testing for all hypoestrogenic and menopausal women. This practice will help start pharmacologic therapy early in the disease, maximally protecting bone and reducing fracture risk. A large-scale randomized prospective trial of bone mineral density testing with long-term follow-up will be needed to crystallize this recommendation.

TABLE 1

Criteria, risk factors for densitometry

The National Osteoporosis Foundation, the North American Menopause Society, the American College of Obstetricians and Gynecologists, and the American Association of Clinical Endocrinologists concur on these criteria and risk factors for bone mineral density testing:
CRITERIA FOR SCREENING
Age 65 years or older
Age less than 65 years with risk factors (see below)
Low-trauma fracture
If densitometry results will influence use of drug treatment
Diseases and treatments associated with osteoporosis (eg, rheumatoid diseases, chronic glucocorticoid therapy)
RISK FACTORS
Prior fracture
Diseases associated with osteoporosis
Body weight less than 127 pounds
Low-trauma fracture in a first-degree relative
Use of chronic glucocorticoid therapy
Cigarette smoking

TABLE 2

Rapid risk assessment: Test bone density if score is 9 or more

The Osteoporosis Risk Assessment Instrument advises testing all women age 65 or older, and menopausal women starting at age 55 who weigh less than 154 pounds and are not taking estrogen.
CRITERIA	POINTS
Age
55-64	5
65-74	9
Older than 74	15
Weight
Less than 60 kg (132 pounds)	9
60 to 70 kg (132 to 154 pounds)	3
Estrogen therapy
Not currently using estrogen	2
Source: Cadarette SM, Jaglal SB, Kreiger N, McIsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ. 2000;162:1289–1294.

When should treatment start?

Fracture begets fracture, treatment reduces risk

Delmas PD, Genant HK, Crans GG, Stock JL, Wong M, Siris E, Adachi JD. Severity of prevalent vertebral fractures and the risk of subsequent vertebral and nonvertebral fractures: results from the MORE trial. Bone. 2003;33:522–532.

 

Baseline fractures in 7,705 postmenopausal women were assessed by semiquantitative spinal radiographs.

In women with severe prevalent vertebral fractures (more than 40% loss of vertebral body height), risk of additional vertebral fractures over 3 years was 38% and risk of new nonvertebral fracture (wrist, hip) was 14%.

Treatment of severe prevalent vertebral fractures with raloxifene 60 mg daily reduced the risk of new vertebral fractures by 26% and new nonvertebral fractures by 47% over 3 years.

To prevent 1 new nonvertebral fracture, 10 patients needed to be treated; to prevent 1 additional nonvertebral fracture, 18 patients needed to be treated.

Women who have already suffered a low-trauma fracture and women with osteoporosis have the greatest risk of fracture. The majority of women who suffer a low-trauma fracture have had neither a bone density measurement nor treatment with a bone medicine, many studies indicate. The Delmas study highlights the importance of prevalent vertebral fractures on the risk of subsequent fractures.

When to treat osteopenia. Many women with osteopenia should be on drug treatment, as should all women with osteoporosis, the NOF advises.

Although a woman with osteoporosis is more likely than a woman with osteopenia to suffer a fracture, the greatest absolute number of fractures occurs in osteopenic women, because that population is so large.

The NOF recommends starting treatment when the T–score measured by dual-energy X-ray absorptiometry (DXA) bone density testing is:

• less than –1.5 and the patient has 1 risk factor, or
  
• less than –2.0.²
  

This approach, it is hoped, will reduce progression to frank osteoporosis and reduce the number of fractures suffered by women with osteopenia.

What are the treatments for low bone mass?

Alendronate, risedronate, and raloxifene

The 3 most commonly used drugs for prevention and treatment of osteoporosis are: 2 bisphosphonates (alendronate and risedronate) and the selective estrogen receptor modulator raloxifene (TABLE 3). All prevent fractures and cost about the same. Alendronate and risedronate are taken by mouth once weekly; raloxifene, daily. A raspberry-flavored liquid alendronate was recently added, for the 10% of women who prefer not to take pills.

Patients must be careful to take alendronate and risedronate in the fasting state and with sufficient water to ensure the pill enters the stomach, then continue to fast another 30 minutes for maximal absorption. The patient needs to remain erect to reduce risk of reflux and esophageal irritation.

Teriparatide

Body JJ, Gaich GA, Scheele WH, et al. A randomized double-blind trial to compare the efficacy of teriparatide with alendronate in postmenopausal women with osteoporosis. J Clin Endocrinol Metab. 2002;87: 4528–4535.

Postmenopausal women with osteoporosis (n = 146) were randomized to receive either teriparatide injections (20 μg daily) plus placebo pills or alendronate 10 mg daily plus placebo injections for a median of 14 months. After 3 months of therapy, bone mineral density in the lumbar spine increased 12.2% and 5.6% in the teriparatide and alendronate groups, respectively. Teriparatide treatment resulted in fewer nonvertebral fractures than alendronate therapy.

Black DM, Greenspan SL, Ensrud KE, et al. The effects of parathyroid hormone and alendronate alone or in combination in postmenopausal osteoporosis. N Engl J Med. 2003;349:1207–1215.

Postmenopausal women with osteoporosis (n = 238) were randomized to receive daily parathyroid hormone (PTH) (full length 1 to 84) injections (100 μg daily), alendronate 10 mg daily, or the combination. After 1 year, lumbar spine bone density as assessed by DXA increased 6.3%, 6.1%, and 4.6% in the PTH alone, PTH plus alendronate and alendronatealone groups, respectively. In this study, PTH plus alendronate conferred no additional benefits over PTH alone.

Recombinant PTH 1-34 (teriparatide, Forteo) was approved in November 2002 for treatment of osteoporosis in postmenopausal women at high risk of fracture. It is very effective and likely superior to alendronate treatment. However, teriparatide, which is an injectable formulation, is expensive and this will likely limit its use to complex cases.

Interestingly, the combination of PTH plus alendronate does not appear to be additive in the treatment of osteoporosis. Given current data, the bisphosphonates should not be combined with PTH.

TABLE 3

Drugs for prevention and treatment of osteoporosis

CLASS, GENERIC NAME, AND INDICATION	BRAND NAME	DOSAGE		APPROXIMATE MONTHLY COST*
		DAILY	WEEKLY
Estrogen for prevention of postmenopausal osteoporosis (loss of bone mass)
Conjugated equine estrogen	Premarin	0.625 mg		$28
Calcitonin-salmon for prevention of progressive loss of bone mass in postmenopausal osteoporosis
Calcitonin	Miacalcin	200 IU by nasal spray		$60
Bisphosphonates for treatment and prevention of osteoporosis in postmenopausal women
Risedronate	Actonel		35 mg	$64
Alendronate	Fosamax		70 mg	$65
Selective estrogen receptor modulator for treatment and prevention of osteoporosis in postmenopausal women
Raloxifene	Evista	60 mg		$71
Parathyroid hormone for treatment of postmenopausal women with osteoporosis who are at high risk for fracture
Teriparatide-PTH 1-35	Forteo	20 μg by injection		$410
* Source: drugstore.com

 

How long should treatment continue?

This question can be answered from both scientific and practical clinical viewpoints.

The practical clinical problem is that when patients ask: “Doctor, how long do I need to take my bone medicine?” they are not emotionally prepared to hear, “Forever.”

It is probably better to say that it’s important to stay on treatment at least 1 to 2 years, and adhere closely to the regimen. When follow-up testing is obtained, the results can influence the next recommendation—which is likely to be that treatment should continue at least 1 or 2 more years. This pattern is more likely to ensure that the patient has high morale and follows the regimen.

Bone mass accrues as treatment continues

Bone HG, Hosking D, Devogelaer JP, et al. Ten years’ experience with alendronate for osteoporosis in postmenopausal women. N Engl J Med. 2004;350:1189–1199.

In postmenopausal women with osteoporosis (n = 86) who were treated for 10 years with alendronate, 10 mg daily, hip and spinal bone density continued to increase throughout follow-up.

Women who stopped therapy gradually lost bone density. After 10 years of alendronate, the increase in bone mineral density was 14% at the lumbar spine and 10% at the hip trochanter.

Studies have demonstrated that bone density continues to increase with up to 10 years of therapy with alendronate.

Alendronate halts bone loss after stopping estrogen

Ascott-Evans BH, Guanabens N, Kivinen S, et al. Alendronate prevents loss of bone density associated with discontinuation of hormone replacement therapy. Arch Intern Med. 2003;163:789–794.

Postmenopausal women (n = 144) who had recently discontinued estrogen therapy were randomized to receive a placebo or alendronate, 10 mg daily. After 1 year, the alendronate group had a 2.3% increase in spinal bone density; the placebo group, a 3.2% decrease.

Once drug therapy stops, bone density begins to decrease—more rapidly after estrogen than after bisphosphonates. In women who stop estrogen, initiation of alendronate blocks bone loss that will otherwise occur. Therefore, women with osteopenia or osteoporosis who stop estrogen therapy should consider starting an alternative bone medicine.

How should you monitor treatment?

Nurses improve adherence

Clowes JA, Peel NFA, Eastell R. The impact of monitoring on adherence and persistence with antiresorptive treatment for postmenopausal osteoporosis: a randomized controlled trial. J Clin Endocrinol Metab. 2004;89:1117–1123.

In this study, 75 postmenopausal women with osteopenia on raloxifene treatment were randomized to 3 different monitoring regimens: no monitoring, nurse interactions with the patient to ensure treatment compliance, or monitoring of urinary markers of bone turnover.

The patients who had the best adherence to therapy had the greatest increase in bone mineral density.

Adherence increased by 57% in the nurse interaction group compared to no monitoring. Measuring urinary markers of bone turnover did not improve adherence or persistence with therapy compared to nurse interactions.

Nurse monitoring of treatment adherence appeared worthwhile, in this comparison of monitoring methods. In routine clinical practice, there is seldom a need to measure markers of bone turnover in women taking bisphosphonates.

Densitometry every 2 years

Greenspan SL, Resnick NM, Parker RA. Combination therapy with hormone replacement and alendronate for prevention of bone loss in elderly women. JAMA. 2003;289: 2525–2533.

A total of 373 women over age 65 were randomized to placebo, conjugated equine estrogen 0.625 mg daily, alendronate 10 mg daily, or both estrogen and alendronate. After 3 years, increases in spinal and hip bone density were greatest in the alendronate plus estrogen group. Alendronate alone was slightly better than estrogen alone in improving bone density at the hip. Alendronate and estrogen were similarly efficacious in improving spine bone density. All active regimens were superior to placebo.

In my practice, I measure bone density every other year to assess response to antiresorptive therapy. Since bone turnover is slow, more frequent measurements are seldom warranted. If bone density stabilizes or increases, I continue therapy.

If bone density decreases significantly on standard monotherapy, I would consider adding estrogen and repeating bone mineral testing in 1 year.

I would also check for secondary causes of bone disease by measuring serum thyroid-stimulating hormone, calcium, albumin, PTH and 25-hydroxyvitamin D.

Alternatively, a woman who has lost bone density on monotherapy can be referred to an endocrinologist.

Dr. Barbieri reports no financial relationships relevant to this article.

Ob/Gyns are well trained to care for more than 99% of women with osteopenia and osteoporosis. Yet most women at risk are not being diagnosed or treated. Since recent publication of randomized clinical trials, however, we have clearer direction on the decisive diagnosis and management questions.

Who should have a bone mass test?

Ob/Gyns know best

Solomon DH, Connelly MT, Rosen CJ, et al. Factors related to the use of bone densitometry: survey responses of 494 primary care physicians in New England. Osteoporosis Int. 2003;14:123–129.

Ob/Gyns were significantly more likely to order bone densitometry than internists and family physicians. However, any physician, including Ob/Gyns, who believed (mistakenly) that calcium-plus-vitamin D effectively treats osteoporosis or that osteoporosis should not be diagnosed by densitometry used screening less frequently than physicians without those beliefs.

Any woman should have bone density testing if it might influence her medical care. Osteoporosis is notoriously difficult to diagnose without densitometry.¹

Besides identifying women at risk of fracture due to osteopenia or osteoporosis who are good candidates for drug therapy, testing can help motivate women to stop smoking, exercise, and take calcium and vitamin D to prevent bone loss. Just as women should know their weight, serum cholesterol, blood pressure, and mammographic findings, hypoestrogenic women should know their T-score.

The issue of relative costs and benefits of bone density testing is complex and continues to evolve. Although national organizations have guidelines (TABLE 1), it is not clear if their sensitivity and specificity are optimal for identifying appropriate candidates for screening.

A reliable, quick tool identifies who to test

Cadarette SM, Jaglal SB, Murray TM, McIsaac WJ, Lawrence L, Brown JP. Evaluation of decision rules for referring women for bone density by dual-energy X-ray absorptiometry. JAMA. 2001;286:57–63.

A 3-item Osteoporosis Risk Assessment Instrument was more sensitive and specific in identifying screening candidates than the National Osteoporosis Foundation (NOF) criteria, according to an analysis of screening algorithms used in 2,365 menopausal women in the Canadian Multicentre Osteoporosis Study. A simple calculation based on age, weight, and estrogen use (TABLE 2) was clinically applicable.

In my practice, I focus on all women who have been hypoestrogenic for 12 to 24 months regardless of age, women with a previous low-trauma fracture, and women who weigh less than 132 pounds. Evidence is mounting that early treatment of bone loss is the best way to prevent future fracture. Well before osteoporosis is detected, significant structural integrity of the spine and hip has been lost.

It is my belief that guidelines will ultimately recommend bone mineral testing for all hypoestrogenic and menopausal women. This practice will help start pharmacologic therapy early in the disease, maximally protecting bone and reducing fracture risk. A large-scale randomized prospective trial of bone mineral density testing with long-term follow-up will be needed to crystallize this recommendation.

TABLE 1

Criteria, risk factors for densitometry

The National Osteoporosis Foundation, the North American Menopause Society, the American College of Obstetricians and Gynecologists, and the American Association of Clinical Endocrinologists concur on these criteria and risk factors for bone mineral density testing:
CRITERIA FOR SCREENING
Age 65 years or older
Age less than 65 years with risk factors (see below)
Low-trauma fracture
If densitometry results will influence use of drug treatment
Diseases and treatments associated with osteoporosis (eg, rheumatoid diseases, chronic glucocorticoid therapy)
RISK FACTORS
Prior fracture
Diseases associated with osteoporosis
Body weight less than 127 pounds
Low-trauma fracture in a first-degree relative
Use of chronic glucocorticoid therapy
Cigarette smoking

TABLE 2

Rapid risk assessment: Test bone density if score is 9 or more

The Osteoporosis Risk Assessment Instrument advises testing all women age 65 or older, and menopausal women starting at age 55 who weigh less than 154 pounds and are not taking estrogen.
CRITERIA	POINTS
Age
55-64	5
65-74	9
Older than 74	15
Weight
Less than 60 kg (132 pounds)	9
60 to 70 kg (132 to 154 pounds)	3
Estrogen therapy
Not currently using estrogen	2
Source: Cadarette SM, Jaglal SB, Kreiger N, McIsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ. 2000;162:1289–1294.

When should treatment start?

Fracture begets fracture, treatment reduces risk

Delmas PD, Genant HK, Crans GG, Stock JL, Wong M, Siris E, Adachi JD. Severity of prevalent vertebral fractures and the risk of subsequent vertebral and nonvertebral fractures: results from the MORE trial. Bone. 2003;33:522–532.

 

Baseline fractures in 7,705 postmenopausal women were assessed by semiquantitative spinal radiographs.

In women with severe prevalent vertebral fractures (more than 40% loss of vertebral body height), risk of additional vertebral fractures over 3 years was 38% and risk of new nonvertebral fracture (wrist, hip) was 14%.

Treatment of severe prevalent vertebral fractures with raloxifene 60 mg daily reduced the risk of new vertebral fractures by 26% and new nonvertebral fractures by 47% over 3 years.

To prevent 1 new nonvertebral fracture, 10 patients needed to be treated; to prevent 1 additional nonvertebral fracture, 18 patients needed to be treated.

Women who have already suffered a low-trauma fracture and women with osteoporosis have the greatest risk of fracture. The majority of women who suffer a low-trauma fracture have had neither a bone density measurement nor treatment with a bone medicine, many studies indicate. The Delmas study highlights the importance of prevalent vertebral fractures on the risk of subsequent fractures.

When to treat osteopenia. Many women with osteopenia should be on drug treatment, as should all women with osteoporosis, the NOF advises.

Although a woman with osteoporosis is more likely than a woman with osteopenia to suffer a fracture, the greatest absolute number of fractures occurs in osteopenic women, because that population is so large.

The NOF recommends starting treatment when the T–score measured by dual-energy X-ray absorptiometry (DXA) bone density testing is:

• less than –1.5 and the patient has 1 risk factor, or
  
• less than –2.0.²
  

This approach, it is hoped, will reduce progression to frank osteoporosis and reduce the number of fractures suffered by women with osteopenia.

What are the treatments for low bone mass?

Alendronate, risedronate, and raloxifene

The 3 most commonly used drugs for prevention and treatment of osteoporosis are: 2 bisphosphonates (alendronate and risedronate) and the selective estrogen receptor modulator raloxifene (TABLE 3). All prevent fractures and cost about the same. Alendronate and risedronate are taken by mouth once weekly; raloxifene, daily. A raspberry-flavored liquid alendronate was recently added, for the 10% of women who prefer not to take pills.

Patients must be careful to take alendronate and risedronate in the fasting state and with sufficient water to ensure the pill enters the stomach, then continue to fast another 30 minutes for maximal absorption. The patient needs to remain erect to reduce risk of reflux and esophageal irritation.

Teriparatide

Body JJ, Gaich GA, Scheele WH, et al. A randomized double-blind trial to compare the efficacy of teriparatide with alendronate in postmenopausal women with osteoporosis. J Clin Endocrinol Metab. 2002;87: 4528–4535.

Postmenopausal women with osteoporosis (n = 146) were randomized to receive either teriparatide injections (20 μg daily) plus placebo pills or alendronate 10 mg daily plus placebo injections for a median of 14 months. After 3 months of therapy, bone mineral density in the lumbar spine increased 12.2% and 5.6% in the teriparatide and alendronate groups, respectively. Teriparatide treatment resulted in fewer nonvertebral fractures than alendronate therapy.

Black DM, Greenspan SL, Ensrud KE, et al. The effects of parathyroid hormone and alendronate alone or in combination in postmenopausal osteoporosis. N Engl J Med. 2003;349:1207–1215.

Postmenopausal women with osteoporosis (n = 238) were randomized to receive daily parathyroid hormone (PTH) (full length 1 to 84) injections (100 μg daily), alendronate 10 mg daily, or the combination. After 1 year, lumbar spine bone density as assessed by DXA increased 6.3%, 6.1%, and 4.6% in the PTH alone, PTH plus alendronate and alendronatealone groups, respectively. In this study, PTH plus alendronate conferred no additional benefits over PTH alone.

Recombinant PTH 1-34 (teriparatide, Forteo) was approved in November 2002 for treatment of osteoporosis in postmenopausal women at high risk of fracture. It is very effective and likely superior to alendronate treatment. However, teriparatide, which is an injectable formulation, is expensive and this will likely limit its use to complex cases.

Interestingly, the combination of PTH plus alendronate does not appear to be additive in the treatment of osteoporosis. Given current data, the bisphosphonates should not be combined with PTH.

TABLE 3

Drugs for prevention and treatment of osteoporosis

CLASS, GENERIC NAME, AND INDICATION	BRAND NAME	DOSAGE		APPROXIMATE MONTHLY COST*
		DAILY	WEEKLY
Estrogen for prevention of postmenopausal osteoporosis (loss of bone mass)
Conjugated equine estrogen	Premarin	0.625 mg		$28
Calcitonin-salmon for prevention of progressive loss of bone mass in postmenopausal osteoporosis
Calcitonin	Miacalcin	200 IU by nasal spray		$60
Bisphosphonates for treatment and prevention of osteoporosis in postmenopausal women
Risedronate	Actonel		35 mg	$64
Alendronate	Fosamax		70 mg	$65
Selective estrogen receptor modulator for treatment and prevention of osteoporosis in postmenopausal women
Raloxifene	Evista	60 mg		$71
Parathyroid hormone for treatment of postmenopausal women with osteoporosis who are at high risk for fracture
Teriparatide-PTH 1-35	Forteo	20 μg by injection		$410
* Source: drugstore.com

 

How long should treatment continue?

This question can be answered from both scientific and practical clinical viewpoints.

The practical clinical problem is that when patients ask: “Doctor, how long do I need to take my bone medicine?” they are not emotionally prepared to hear, “Forever.”

It is probably better to say that it’s important to stay on treatment at least 1 to 2 years, and adhere closely to the regimen. When follow-up testing is obtained, the results can influence the next recommendation—which is likely to be that treatment should continue at least 1 or 2 more years. This pattern is more likely to ensure that the patient has high morale and follows the regimen.

Bone mass accrues as treatment continues

Bone HG, Hosking D, Devogelaer JP, et al. Ten years’ experience with alendronate for osteoporosis in postmenopausal women. N Engl J Med. 2004;350:1189–1199.

In postmenopausal women with osteoporosis (n = 86) who were treated for 10 years with alendronate, 10 mg daily, hip and spinal bone density continued to increase throughout follow-up.

Women who stopped therapy gradually lost bone density. After 10 years of alendronate, the increase in bone mineral density was 14% at the lumbar spine and 10% at the hip trochanter.

Studies have demonstrated that bone density continues to increase with up to 10 years of therapy with alendronate.

Alendronate halts bone loss after stopping estrogen

Ascott-Evans BH, Guanabens N, Kivinen S, et al. Alendronate prevents loss of bone density associated with discontinuation of hormone replacement therapy. Arch Intern Med. 2003;163:789–794.

Postmenopausal women (n = 144) who had recently discontinued estrogen therapy were randomized to receive a placebo or alendronate, 10 mg daily. After 1 year, the alendronate group had a 2.3% increase in spinal bone density; the placebo group, a 3.2% decrease.

Once drug therapy stops, bone density begins to decrease—more rapidly after estrogen than after bisphosphonates. In women who stop estrogen, initiation of alendronate blocks bone loss that will otherwise occur. Therefore, women with osteopenia or osteoporosis who stop estrogen therapy should consider starting an alternative bone medicine.

How should you monitor treatment?

Nurses improve adherence

Clowes JA, Peel NFA, Eastell R. The impact of monitoring on adherence and persistence with antiresorptive treatment for postmenopausal osteoporosis: a randomized controlled trial. J Clin Endocrinol Metab. 2004;89:1117–1123.

In this study, 75 postmenopausal women with osteopenia on raloxifene treatment were randomized to 3 different monitoring regimens: no monitoring, nurse interactions with the patient to ensure treatment compliance, or monitoring of urinary markers of bone turnover.

The patients who had the best adherence to therapy had the greatest increase in bone mineral density.

Adherence increased by 57% in the nurse interaction group compared to no monitoring. Measuring urinary markers of bone turnover did not improve adherence or persistence with therapy compared to nurse interactions.

Nurse monitoring of treatment adherence appeared worthwhile, in this comparison of monitoring methods. In routine clinical practice, there is seldom a need to measure markers of bone turnover in women taking bisphosphonates.

Densitometry every 2 years

Greenspan SL, Resnick NM, Parker RA. Combination therapy with hormone replacement and alendronate for prevention of bone loss in elderly women. JAMA. 2003;289: 2525–2533.

A total of 373 women over age 65 were randomized to placebo, conjugated equine estrogen 0.625 mg daily, alendronate 10 mg daily, or both estrogen and alendronate. After 3 years, increases in spinal and hip bone density were greatest in the alendronate plus estrogen group. Alendronate alone was slightly better than estrogen alone in improving bone density at the hip. Alendronate and estrogen were similarly efficacious in improving spine bone density. All active regimens were superior to placebo.

In my practice, I measure bone density every other year to assess response to antiresorptive therapy. Since bone turnover is slow, more frequent measurements are seldom warranted. If bone density stabilizes or increases, I continue therapy.

If bone density decreases significantly on standard monotherapy, I would consider adding estrogen and repeating bone mineral testing in 1 year.

I would also check for secondary causes of bone disease by measuring serum thyroid-stimulating hormone, calcium, albumin, PTH and 25-hydroxyvitamin D.

Alternatively, a woman who has lost bone density on monotherapy can be referred to an endocrinologist.

Dr. Barbieri reports no financial relationships relevant to this article.

Ob/Gyns are well trained to care for more than 99% of women with osteopenia and osteoporosis. Yet most women at risk are not being diagnosed or treated. Since recent publication of randomized clinical trials, however, we have clearer direction on the decisive diagnosis and management questions.

Who should have a bone mass test?

Ob/Gyns know best

Solomon DH, Connelly MT, Rosen CJ, et al. Factors related to the use of bone densitometry: survey responses of 494 primary care physicians in New England. Osteoporosis Int. 2003;14:123–129.

Ob/Gyns were significantly more likely to order bone densitometry than internists and family physicians. However, any physician, including Ob/Gyns, who believed (mistakenly) that calcium-plus-vitamin D effectively treats osteoporosis or that osteoporosis should not be diagnosed by densitometry used screening less frequently than physicians without those beliefs.

Any woman should have bone density testing if it might influence her medical care. Osteoporosis is notoriously difficult to diagnose without densitometry.¹

Besides identifying women at risk of fracture due to osteopenia or osteoporosis who are good candidates for drug therapy, testing can help motivate women to stop smoking, exercise, and take calcium and vitamin D to prevent bone loss. Just as women should know their weight, serum cholesterol, blood pressure, and mammographic findings, hypoestrogenic women should know their T-score.

The issue of relative costs and benefits of bone density testing is complex and continues to evolve. Although national organizations have guidelines (TABLE 1), it is not clear if their sensitivity and specificity are optimal for identifying appropriate candidates for screening.

A reliable, quick tool identifies who to test

Cadarette SM, Jaglal SB, Murray TM, McIsaac WJ, Lawrence L, Brown JP. Evaluation of decision rules for referring women for bone density by dual-energy X-ray absorptiometry. JAMA. 2001;286:57–63.

A 3-item Osteoporosis Risk Assessment Instrument was more sensitive and specific in identifying screening candidates than the National Osteoporosis Foundation (NOF) criteria, according to an analysis of screening algorithms used in 2,365 menopausal women in the Canadian Multicentre Osteoporosis Study. A simple calculation based on age, weight, and estrogen use (TABLE 2) was clinically applicable.

In my practice, I focus on all women who have been hypoestrogenic for 12 to 24 months regardless of age, women with a previous low-trauma fracture, and women who weigh less than 132 pounds. Evidence is mounting that early treatment of bone loss is the best way to prevent future fracture. Well before osteoporosis is detected, significant structural integrity of the spine and hip has been lost.

It is my belief that guidelines will ultimately recommend bone mineral testing for all hypoestrogenic and menopausal women. This practice will help start pharmacologic therapy early in the disease, maximally protecting bone and reducing fracture risk. A large-scale randomized prospective trial of bone mineral density testing with long-term follow-up will be needed to crystallize this recommendation.

TABLE 1

Criteria, risk factors for densitometry

The National Osteoporosis Foundation, the North American Menopause Society, the American College of Obstetricians and Gynecologists, and the American Association of Clinical Endocrinologists concur on these criteria and risk factors for bone mineral density testing:
CRITERIA FOR SCREENING
Age 65 years or older
Age less than 65 years with risk factors (see below)
Low-trauma fracture
If densitometry results will influence use of drug treatment
Diseases and treatments associated with osteoporosis (eg, rheumatoid diseases, chronic glucocorticoid therapy)
RISK FACTORS
Prior fracture
Diseases associated with osteoporosis
Body weight less than 127 pounds
Low-trauma fracture in a first-degree relative
Use of chronic glucocorticoid therapy
Cigarette smoking

TABLE 2

Rapid risk assessment: Test bone density if score is 9 or more

The Osteoporosis Risk Assessment Instrument advises testing all women age 65 or older, and menopausal women starting at age 55 who weigh less than 154 pounds and are not taking estrogen.
CRITERIA	POINTS
Age
55-64	5
65-74	9
Older than 74	15
Weight
Less than 60 kg (132 pounds)	9
60 to 70 kg (132 to 154 pounds)	3
Estrogen therapy
Not currently using estrogen	2
Source: Cadarette SM, Jaglal SB, Kreiger N, McIsaac WJ, Darlington GA, Tu JV. Development and validation of the Osteoporosis Risk Assessment Instrument to facilitate selection of women for bone densitometry. CMAJ. 2000;162:1289–1294.

When should treatment start?

Fracture begets fracture, treatment reduces risk

Delmas PD, Genant HK, Crans GG, Stock JL, Wong M, Siris E, Adachi JD. Severity of prevalent vertebral fractures and the risk of subsequent vertebral and nonvertebral fractures: results from the MORE trial. Bone. 2003;33:522–532.

 

Baseline fractures in 7,705 postmenopausal women were assessed by semiquantitative spinal radiographs.

In women with severe prevalent vertebral fractures (more than 40% loss of vertebral body height), risk of additional vertebral fractures over 3 years was 38% and risk of new nonvertebral fracture (wrist, hip) was 14%.

Treatment of severe prevalent vertebral fractures with raloxifene 60 mg daily reduced the risk of new vertebral fractures by 26% and new nonvertebral fractures by 47% over 3 years.

To prevent 1 new nonvertebral fracture, 10 patients needed to be treated; to prevent 1 additional nonvertebral fracture, 18 patients needed to be treated.

Women who have already suffered a low-trauma fracture and women with osteoporosis have the greatest risk of fracture. The majority of women who suffer a low-trauma fracture have had neither a bone density measurement nor treatment with a bone medicine, many studies indicate. The Delmas study highlights the importance of prevalent vertebral fractures on the risk of subsequent fractures.

When to treat osteopenia. Many women with osteopenia should be on drug treatment, as should all women with osteoporosis, the NOF advises.

Although a woman with osteoporosis is more likely than a woman with osteopenia to suffer a fracture, the greatest absolute number of fractures occurs in osteopenic women, because that population is so large.

The NOF recommends starting treatment when the T–score measured by dual-energy X-ray absorptiometry (DXA) bone density testing is:

• less than –1.5 and the patient has 1 risk factor, or
  
• less than –2.0.²
  

This approach, it is hoped, will reduce progression to frank osteoporosis and reduce the number of fractures suffered by women with osteopenia.

What are the treatments for low bone mass?

Alendronate, risedronate, and raloxifene

The 3 most commonly used drugs for prevention and treatment of osteoporosis are: 2 bisphosphonates (alendronate and risedronate) and the selective estrogen receptor modulator raloxifene (TABLE 3). All prevent fractures and cost about the same. Alendronate and risedronate are taken by mouth once weekly; raloxifene, daily. A raspberry-flavored liquid alendronate was recently added, for the 10% of women who prefer not to take pills.

Patients must be careful to take alendronate and risedronate in the fasting state and with sufficient water to ensure the pill enters the stomach, then continue to fast another 30 minutes for maximal absorption. The patient needs to remain erect to reduce risk of reflux and esophageal irritation.

Teriparatide

Body JJ, Gaich GA, Scheele WH, et al. A randomized double-blind trial to compare the efficacy of teriparatide with alendronate in postmenopausal women with osteoporosis. J Clin Endocrinol Metab. 2002;87: 4528–4535.

Postmenopausal women with osteoporosis (n = 146) were randomized to receive either teriparatide injections (20 μg daily) plus placebo pills or alendronate 10 mg daily plus placebo injections for a median of 14 months. After 3 months of therapy, bone mineral density in the lumbar spine increased 12.2% and 5.6% in the teriparatide and alendronate groups, respectively. Teriparatide treatment resulted in fewer nonvertebral fractures than alendronate therapy.

Black DM, Greenspan SL, Ensrud KE, et al. The effects of parathyroid hormone and alendronate alone or in combination in postmenopausal osteoporosis. N Engl J Med. 2003;349:1207–1215.

Postmenopausal women with osteoporosis (n = 238) were randomized to receive daily parathyroid hormone (PTH) (full length 1 to 84) injections (100 μg daily), alendronate 10 mg daily, or the combination. After 1 year, lumbar spine bone density as assessed by DXA increased 6.3%, 6.1%, and 4.6% in the PTH alone, PTH plus alendronate and alendronatealone groups, respectively. In this study, PTH plus alendronate conferred no additional benefits over PTH alone.

Recombinant PTH 1-34 (teriparatide, Forteo) was approved in November 2002 for treatment of osteoporosis in postmenopausal women at high risk of fracture. It is very effective and likely superior to alendronate treatment. However, teriparatide, which is an injectable formulation, is expensive and this will likely limit its use to complex cases.

Interestingly, the combination of PTH plus alendronate does not appear to be additive in the treatment of osteoporosis. Given current data, the bisphosphonates should not be combined with PTH.

TABLE 3

Drugs for prevention and treatment of osteoporosis

CLASS, GENERIC NAME, AND INDICATION	BRAND NAME	DOSAGE		APPROXIMATE MONTHLY COST*
		DAILY	WEEKLY
Estrogen for prevention of postmenopausal osteoporosis (loss of bone mass)
Conjugated equine estrogen	Premarin	0.625 mg		$28
Calcitonin-salmon for prevention of progressive loss of bone mass in postmenopausal osteoporosis
Calcitonin	Miacalcin	200 IU by nasal spray		$60
Bisphosphonates for treatment and prevention of osteoporosis in postmenopausal women
Risedronate	Actonel		35 mg	$64
Alendronate	Fosamax		70 mg	$65
Selective estrogen receptor modulator for treatment and prevention of osteoporosis in postmenopausal women
Raloxifene	Evista	60 mg		$71
Parathyroid hormone for treatment of postmenopausal women with osteoporosis who are at high risk for fracture
Teriparatide-PTH 1-35	Forteo	20 μg by injection		$410
* Source: drugstore.com

 

How long should treatment continue?

This question can be answered from both scientific and practical clinical viewpoints.

The practical clinical problem is that when patients ask: “Doctor, how long do I need to take my bone medicine?” they are not emotionally prepared to hear, “Forever.”

It is probably better to say that it’s important to stay on treatment at least 1 to 2 years, and adhere closely to the regimen. When follow-up testing is obtained, the results can influence the next recommendation—which is likely to be that treatment should continue at least 1 or 2 more years. This pattern is more likely to ensure that the patient has high morale and follows the regimen.

Bone mass accrues as treatment continues

Bone HG, Hosking D, Devogelaer JP, et al. Ten years’ experience with alendronate for osteoporosis in postmenopausal women. N Engl J Med. 2004;350:1189–1199.

In postmenopausal women with osteoporosis (n = 86) who were treated for 10 years with alendronate, 10 mg daily, hip and spinal bone density continued to increase throughout follow-up.

Women who stopped therapy gradually lost bone density. After 10 years of alendronate, the increase in bone mineral density was 14% at the lumbar spine and 10% at the hip trochanter.

Studies have demonstrated that bone density continues to increase with up to 10 years of therapy with alendronate.

Alendronate halts bone loss after stopping estrogen

Ascott-Evans BH, Guanabens N, Kivinen S, et al. Alendronate prevents loss of bone density associated with discontinuation of hormone replacement therapy. Arch Intern Med. 2003;163:789–794.

Postmenopausal women (n = 144) who had recently discontinued estrogen therapy were randomized to receive a placebo or alendronate, 10 mg daily. After 1 year, the alendronate group had a 2.3% increase in spinal bone density; the placebo group, a 3.2% decrease.

Once drug therapy stops, bone density begins to decrease—more rapidly after estrogen than after bisphosphonates. In women who stop estrogen, initiation of alendronate blocks bone loss that will otherwise occur. Therefore, women with osteopenia or osteoporosis who stop estrogen therapy should consider starting an alternative bone medicine.

How should you monitor treatment?

Nurses improve adherence

Clowes JA, Peel NFA, Eastell R. The impact of monitoring on adherence and persistence with antiresorptive treatment for postmenopausal osteoporosis: a randomized controlled trial. J Clin Endocrinol Metab. 2004;89:1117–1123.

In this study, 75 postmenopausal women with osteopenia on raloxifene treatment were randomized to 3 different monitoring regimens: no monitoring, nurse interactions with the patient to ensure treatment compliance, or monitoring of urinary markers of bone turnover.

The patients who had the best adherence to therapy had the greatest increase in bone mineral density.

Adherence increased by 57% in the nurse interaction group compared to no monitoring. Measuring urinary markers of bone turnover did not improve adherence or persistence with therapy compared to nurse interactions.

Nurse monitoring of treatment adherence appeared worthwhile, in this comparison of monitoring methods. In routine clinical practice, there is seldom a need to measure markers of bone turnover in women taking bisphosphonates.

Densitometry every 2 years

Greenspan SL, Resnick NM, Parker RA. Combination therapy with hormone replacement and alendronate for prevention of bone loss in elderly women. JAMA. 2003;289: 2525–2533.

A total of 373 women over age 65 were randomized to placebo, conjugated equine estrogen 0.625 mg daily, alendronate 10 mg daily, or both estrogen and alendronate. After 3 years, increases in spinal and hip bone density were greatest in the alendronate plus estrogen group. Alendronate alone was slightly better than estrogen alone in improving bone density at the hip. Alendronate and estrogen were similarly efficacious in improving spine bone density. All active regimens were superior to placebo.

In my practice, I measure bone density every other year to assess response to antiresorptive therapy. Since bone turnover is slow, more frequent measurements are seldom warranted. If bone density stabilizes or increases, I continue therapy.

If bone density decreases significantly on standard monotherapy, I would consider adding estrogen and repeating bone mineral testing in 1 year.

I would also check for secondary causes of bone disease by measuring serum thyroid-stimulating hormone, calcium, albumin, PTH and 25-hydroxyvitamin D.

Alternatively, a woman who has lost bone density on monotherapy can be referred to an endocrinologist.

Dr. Barbieri reports no financial relationships relevant to this article.

THE CASE: SIGNS AND SYMPTOMS

A 29–year–old nullipara at 18 weeks’ gestation complains of fevers and back pain. She had a diagnosis of urinary tract infection with sulfonamide-resistant Escherichia coli at 9 weeks of gestation, which was treated with nitrofurantoin, 100 mg by mouth twice a day for 7 days. A test of cure by urine culture was negative.

Now her temperature is 101°F and she has right costovertebral angle tenderness.

How should you proceed?

Anatomy is destiny, in the case of susceptibility to urinary tract infection (UTI). The female urethra is only 3 cm to 4 cm long, and its proximity to the vagina, anus, and rectum facilitates colonization of normal gastrointestinal flora in the bladder.¹

Sexual activity also facilitates migration of normal gastrointestinal flora to the female urethra.²

Anatomical features of pregnancy exacerbate the female predisposition to urinary tract infection. In pregnancy, hormonal and mechanical changes that occur in the urinary tract lead to urinary stasis and ureterovesical reflux—setting the stage for urinary tract infection (FIGURE 1).

Who should be screened?

All pregnant women should be screened for UTI early in pregnancy, according to the American College of Obstetricians and Gynecologists.³

I recommend a urine culture screening for all pregnant women at their first prenatal visit.

Screen often if she has risk factors

I recommend frequent screening (at least every trimester) by urine culture, in pregnant women with any of these risk factors:

• diabetes mellitus, including gestational diabetes⁴;
  
• urologic abnormalities—specifically, neurogenic bladder;
  
• prepregnancy (for example, 2 to 3 infections per year) and antepartum history of UTI prior to initiation of prenatal care⁵;
  
• sickle cell hemoglobinopathy.⁵
  

Which test is best?

The gold standard for detecting bacteria in urine is by culture.

Which threshold to use?

The standard definition of a positive urine culture from a clean-catch, midstream, voided specimen is ≥100,000 colony forming units (CFU) per mL of a single organism. However, in symptomatic patients, the test’s sensitivity is increased by lowering the cut-off to 100 CFU/mL of a single organism.⁶ In women with urinary symptoms, only 50% of patients had 100,000 CFU/mL by urine culture collected from clean-catch, midstream, voided specimens, though all of them had positive cultures from suprapubic taps.

The clean-catch, midstream, voided specimen is the specimen of choice for practical purposes, since it is noninvasive and easily obtained in the office setting.

For the record: The presence of any organism represents UTI in specimens obtained via suprapubic aspiration of the bladder; 100 CFU/mL of a single organism is positive for specimens obtained by urethral catheterization.

I recommend that, when obtaining urine cultures via clean-catch, midstream, voided specimens:

• for asymptomatic patients, use ≥100,000 CFU/mL of a single organism.
  
• in symptomatic patients, use ≥100 CFU/mL of a single organism.
  

What about rapid tests?

Urinary sediment analysis and urine dipstick testing offer speed and low cost, but with lower accuracy than urine cultures, which require 24 to 48 hours for results and cost more.

Urinary sediment analysis can diagnose pyuria, defined as a clean-catch, midstream, voided specimen, which is spun and which has >10 leukocytes per high-power field.

Pyuria can occur without infection due to:

• previous treatment with antibiotics,
  
• contamination of urine sample by sterilizing solution,
  
• contamination of urine sample with vaginal leukocytes,
  
• chronic interstitial nephritis (such as analgesic abuse),
  
• uroepithelial tumor, and
  
• nephrolithiasis.
  

Pyuria on urinalysis has low sensitivity (25%) but high specificity (99%).

Bacteria visualized on microscopic examination is more sensitive (75%) but less specific (60%).⁷

Urinary dipstick testing—fast, convenient, and low in cost—is considered positive if it identifies either leukocyte esterase or nitrite. Positive leukocyte esterase signifies pyuria. Positive nitrite indicates the presence of enteric organisms that convert urinary nitrate to nitrite.

With either finding, dipstick sensitivity is only 50%, although specificity is 97%.⁷

I recommend:

• If a symptomatic patient’s rapid test is positive, obtain a urine culture, empirically treat for UTI, and then use urine culture results to decide whether to continue treatment.
  
• If an asymptomatic patient’s rapid test is positive, obtain a urine culture and treat only if the culture is positive.
  

What urinary tract disorders occur in pregnancy?

First, determine if the patient has urinary tract symptoms and, if so, whether the symptoms are typical of upper or lower urinary tract infections.

 

Lower urinary tract symptoms:

• dysuria
  
• frequency
  
• urgency
  
• suprapubic pain
  
• hematuria in the absence of fever and systemic symptoms
  

Frequency, of course, is difficult to ascertain in pregnancy, since most women experience this complication secondary to increased urinary output due to higher fluid intake, increased plasma volume expansion, and increases in renal blood flow and glomerular filtration rate.

Upper urinary tract symptoms:

• fever
  
• chills
  
• flank pain
  
• nausea and vomiting
  
• The patient may or may not have the symptoms of lower urinary tract infection, as well.
  

Positive culture and no symptoms

This profile is typical of asymptomatic bacteriuria, a lower urinary tract infection that occurs in 2% to 7% of pregnancies.¹

Positive culture with symptoms

This profile probably reflects either:

• Acute cystitis, a lower urinary tract infection affecting 1% to 2% of pregnancies,⁸ or
  
• Acute pyelonephritis, an upper urinary tract infection affecting 2% of pregnancies.⁹
  

What are the consequences of UTI in pregnancy?

Maternal complications

Asymptomatic bacteriuria does not plague the patient with bothersome effects, but if left untreated, asymptomatic bacteriuria will progress to symptomatic UTI: 25% will develop acute pyelonephritis, compared to 3% to 4% of treated patients¹⁰; 20% of women with severe pyelonephritis develop serious complications,¹¹ including:

• sepsis and septic shock,
  
• hemolysis and thrombocytopenia,¹²
  
• acute respiratory distress syndrome,¹³
  
• renal insufficiency.¹⁴
  

Adverse fetal outcomes

Untreated asymptomatic bacteriuria is associated with preterm delivery and low birthweight.^15-17

Acute pyelonephritis is linked to preterm birth.^18,19 Kaul et al,²⁰ in an experimental model of pyelonephritis in mice, confirmed that E. coli plays an important role in the pathogenesis of preterm delivery and low birthweight.

What is the best treatment regimen?

Data are insufficient to recommend any specific regimen.^21,22 The following strategies are based on evaluation of review articles.^3,23

Asymptomatic bacteriuria and acute cystitis

Nitrofurantoin monohydrate macrocrystals is my first-line treatment. Nitrofurantoin has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

If nitrofurantoin is not effective, I change antibiotics based on urine culture antibiotic sensitivity profiles (FIGURE 2).

Keep in mind the current resistance of E. coli to antibiotics: ampicillin, 28% to 39%; trimethoprim-sulfamethoxazole, 31%; and first-generation cephalosporins, 9% to 19%.²⁴

Single-dose treatment for pregnant women with asymptomatic bacteriuria has been evaluated, given its lower cost and better compliance. However, evidence is insufficient to determine whether single-dose or longer-duration regimens are more effective.²⁵

I recommend longer-duration dosages for now, until a large randomized controlled trial can derive conclusive data.

Remember that treatment success is not contingent upon duration of therapy—just be sure that the test-of-cure urine culture is negative 1 to 2 weeks after treatment is completed.

Acute pyelonephritis

Management should include the following:

• hospitalization
  
• urine and blood cultures
  
• laboratory studies of complete blood cell count, electrolytes, creatinine, and liver function
  
• monitoring of vital signs and urine output
  
• intravenous (IV) crystalloid fluid to maintain urine output
  
• IV antibiotics (FIGURE 3).
  

If symptoms persist after 48 hours despite adequate IV antibiotic therapy, consider other causes such as resistant organisms, nephrolithiasis, perinephric abscesses, renal obstruction, or other infections.

Consider imaging by renal ultrasound to assess the presence of nephrolithiasis, perinephric abscess, or obstruction.

I recommend inpatient treatment for pregnant women with acute pyelonephritis at this time, until further studies are available.

To evaluate outpatient treatment, Millar and colleagues randomized 120 women under 24 weeks’ gestation either to inpatient IV cefazolin until 48 hours afebrile or to outpatient ceftriaxone intramuscularly. (Both treatment arms completed a course of oral cephalexin.) There were no differences in therapeutic response or birth outcomes, but 6 patients in the outpatient arm required hospitalization for IV therapy and 1 woman developed sepsis.²⁶

The same researchers studied 92 patients of more than 24 weeks’ gestation who received 2 doses of ceftriaxone intramuscularly, then were randomized to either continued inpatient therapy until 48 hours afebrile or discharge with reevaluation as an outpatient in 48 to 72 hours. Again, there were no differences in therapeutic response or birth outcomes; however, almost two-thirds of patients were excluded from the study as they did not meet criteria for outpatient management, due to sepsis, preterm labor, or concurrent medical conditions.²⁷

Adequate antibiotic coverage is crucial

To ensure adequate antibiotic coverage when treating UTI, it is important to understand which organisms cause these infections in pregnancy.

E. coli causes 75% to 90% of UTIs in nonpregnant women.²⁸Staphylococcus saprophyticus causes 10% to 15% of UTIs in nonpregnant women, but less in pregnant women.

Group B Streptococcus (GBS)—another gram-positive organism—has important implications for pregnant women: Intrapartum prophylaxis is important, to prevent neonatal GBS disease.²⁹

Klebsiella, Enterobacter, Proteus, and Enterococcus species²⁸ infrequently cause UTI in pregnancy.

What about prophylaxis and follow-up cultures?

 

Expect recurrence

One third of pregnant women diagnosed with UTI will have recurrence.¹ Recurrence is either relapse (same strain, within 2 weeks of completing initial treatment for the original infection) or reinfection (different strain or same strain after more than 2 weeks).

2 UTIs or pyelonephritis warrant suppressive therapy

I recommend suppressive therapy if a pregnant woman is diagnosed with 2 lower urinary tract infections or acute pyelonephritis (TABLE).

Nitrofurantoin is the preferred agent, as it has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

Start only after eradication of the acute infection, as evidenced by a negative test-of-cure urine culture at least 1 to 2 weeks after treatment is discontinued.²³

Monthly urine cultures until delivery

I recommend monthly follow-up urine cultures until delivery.

Periodic follow-up screening is often recommended, but opinions differ on which test to use or how often to screen.

THE CASE: DIAGNOSIS, TREATMENT, FOLLOW-UP, AND OUTCOME

The patient with upper urinary tract symptoms had a white blood cell count of 15, a urine dipstick positive for leukocyte esterase and nitrites, and a urine sediment analysis indicating pyuria.

She was diagnosed with acute pyelonephritis and started on ampicillin and gentamicin intravenously. Her urine culture drawn upon admission grew >100,000 CFU/mL of sulfonamide-resistant E. coli.

Within 48 hours, she showed clinical improvement and was discharged home with a 10-day course of nitrofurantoin. One week after completing treatment, her test-of-cure urine culture was negative and she was started on nitrofurantoin 50 mg every night at bedtime.

For the rest of pregnancy, she underwent monthly screening urine cultures, which remained negative. She had an uncomplicated delivery at 38 weeks of gestation.

TABLE

Suppressive therapy to prevent UTI recurrence

Suppressive therapy is recommended for any pregnant woman with:
2 lower urinary tract infections or acute pyelonephritis.
Do not initiate suppressive therapy until a negative test-of-cure urine culture confirms eradication of the acute infection.
ANTIBIOTIC	DOSE (ORAL)
Nitrofurantoin monohydrate macrocrystals	50 mg at bedtime
or
Cephalexin	250 mg at bedtime

Dr. Chen reports support from a Women’s Reproductive Health Research Career Development Center grant from the National Institutes of Child Health and Human Development.

THE CASE: SIGNS AND SYMPTOMS

A 29–year–old nullipara at 18 weeks’ gestation complains of fevers and back pain. She had a diagnosis of urinary tract infection with sulfonamide-resistant Escherichia coli at 9 weeks of gestation, which was treated with nitrofurantoin, 100 mg by mouth twice a day for 7 days. A test of cure by urine culture was negative.

Now her temperature is 101°F and she has right costovertebral angle tenderness.

How should you proceed?

Anatomy is destiny, in the case of susceptibility to urinary tract infection (UTI). The female urethra is only 3 cm to 4 cm long, and its proximity to the vagina, anus, and rectum facilitates colonization of normal gastrointestinal flora in the bladder.¹

Sexual activity also facilitates migration of normal gastrointestinal flora to the female urethra.²

Anatomical features of pregnancy exacerbate the female predisposition to urinary tract infection. In pregnancy, hormonal and mechanical changes that occur in the urinary tract lead to urinary stasis and ureterovesical reflux—setting the stage for urinary tract infection (FIGURE 1).

Who should be screened?

All pregnant women should be screened for UTI early in pregnancy, according to the American College of Obstetricians and Gynecologists.³

I recommend a urine culture screening for all pregnant women at their first prenatal visit.

Screen often if she has risk factors

I recommend frequent screening (at least every trimester) by urine culture, in pregnant women with any of these risk factors:

• diabetes mellitus, including gestational diabetes⁴;
  
• urologic abnormalities—specifically, neurogenic bladder;
  
• prepregnancy (for example, 2 to 3 infections per year) and antepartum history of UTI prior to initiation of prenatal care⁵;
  
• sickle cell hemoglobinopathy.⁵
  

Which test is best?

The gold standard for detecting bacteria in urine is by culture.

Which threshold to use?

The standard definition of a positive urine culture from a clean-catch, midstream, voided specimen is ≥100,000 colony forming units (CFU) per mL of a single organism. However, in symptomatic patients, the test’s sensitivity is increased by lowering the cut-off to 100 CFU/mL of a single organism.⁶ In women with urinary symptoms, only 50% of patients had 100,000 CFU/mL by urine culture collected from clean-catch, midstream, voided specimens, though all of them had positive cultures from suprapubic taps.

The clean-catch, midstream, voided specimen is the specimen of choice for practical purposes, since it is noninvasive and easily obtained in the office setting.

For the record: The presence of any organism represents UTI in specimens obtained via suprapubic aspiration of the bladder; 100 CFU/mL of a single organism is positive for specimens obtained by urethral catheterization.

I recommend that, when obtaining urine cultures via clean-catch, midstream, voided specimens:

• for asymptomatic patients, use ≥100,000 CFU/mL of a single organism.
  
• in symptomatic patients, use ≥100 CFU/mL of a single organism.
  

What about rapid tests?

Urinary sediment analysis and urine dipstick testing offer speed and low cost, but with lower accuracy than urine cultures, which require 24 to 48 hours for results and cost more.

Urinary sediment analysis can diagnose pyuria, defined as a clean-catch, midstream, voided specimen, which is spun and which has >10 leukocytes per high-power field.

Pyuria can occur without infection due to:

• previous treatment with antibiotics,
  
• contamination of urine sample by sterilizing solution,
  
• contamination of urine sample with vaginal leukocytes,
  
• chronic interstitial nephritis (such as analgesic abuse),
  
• uroepithelial tumor, and
  
• nephrolithiasis.
  

Pyuria on urinalysis has low sensitivity (25%) but high specificity (99%).

Bacteria visualized on microscopic examination is more sensitive (75%) but less specific (60%).⁷

Urinary dipstick testing—fast, convenient, and low in cost—is considered positive if it identifies either leukocyte esterase or nitrite. Positive leukocyte esterase signifies pyuria. Positive nitrite indicates the presence of enteric organisms that convert urinary nitrate to nitrite.

With either finding, dipstick sensitivity is only 50%, although specificity is 97%.⁷

I recommend:

• If a symptomatic patient’s rapid test is positive, obtain a urine culture, empirically treat for UTI, and then use urine culture results to decide whether to continue treatment.
  
• If an asymptomatic patient’s rapid test is positive, obtain a urine culture and treat only if the culture is positive.
  

What urinary tract disorders occur in pregnancy?

First, determine if the patient has urinary tract symptoms and, if so, whether the symptoms are typical of upper or lower urinary tract infections.

 

Lower urinary tract symptoms:

• dysuria
  
• frequency
  
• urgency
  
• suprapubic pain
  
• hematuria in the absence of fever and systemic symptoms
  

Frequency, of course, is difficult to ascertain in pregnancy, since most women experience this complication secondary to increased urinary output due to higher fluid intake, increased plasma volume expansion, and increases in renal blood flow and glomerular filtration rate.

Upper urinary tract symptoms:

• fever
  
• chills
  
• flank pain
  
• nausea and vomiting
  
• The patient may or may not have the symptoms of lower urinary tract infection, as well.
  

Positive culture and no symptoms

This profile is typical of asymptomatic bacteriuria, a lower urinary tract infection that occurs in 2% to 7% of pregnancies.¹

Positive culture with symptoms

This profile probably reflects either:

• Acute cystitis, a lower urinary tract infection affecting 1% to 2% of pregnancies,⁸ or
  
• Acute pyelonephritis, an upper urinary tract infection affecting 2% of pregnancies.⁹
  

What are the consequences of UTI in pregnancy?

Maternal complications

Asymptomatic bacteriuria does not plague the patient with bothersome effects, but if left untreated, asymptomatic bacteriuria will progress to symptomatic UTI: 25% will develop acute pyelonephritis, compared to 3% to 4% of treated patients¹⁰; 20% of women with severe pyelonephritis develop serious complications,¹¹ including:

• sepsis and septic shock,
  
• hemolysis and thrombocytopenia,¹²
  
• acute respiratory distress syndrome,¹³
  
• renal insufficiency.¹⁴
  

Adverse fetal outcomes

Untreated asymptomatic bacteriuria is associated with preterm delivery and low birthweight.^15-17

Acute pyelonephritis is linked to preterm birth.^18,19 Kaul et al,²⁰ in an experimental model of pyelonephritis in mice, confirmed that E. coli plays an important role in the pathogenesis of preterm delivery and low birthweight.

What is the best treatment regimen?

Data are insufficient to recommend any specific regimen.^21,22 The following strategies are based on evaluation of review articles.^3,23

Asymptomatic bacteriuria and acute cystitis

Nitrofurantoin monohydrate macrocrystals is my first-line treatment. Nitrofurantoin has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

If nitrofurantoin is not effective, I change antibiotics based on urine culture antibiotic sensitivity profiles (FIGURE 2).

Keep in mind the current resistance of E. coli to antibiotics: ampicillin, 28% to 39%; trimethoprim-sulfamethoxazole, 31%; and first-generation cephalosporins, 9% to 19%.²⁴

Single-dose treatment for pregnant women with asymptomatic bacteriuria has been evaluated, given its lower cost and better compliance. However, evidence is insufficient to determine whether single-dose or longer-duration regimens are more effective.²⁵

I recommend longer-duration dosages for now, until a large randomized controlled trial can derive conclusive data.

Remember that treatment success is not contingent upon duration of therapy—just be sure that the test-of-cure urine culture is negative 1 to 2 weeks after treatment is completed.

Acute pyelonephritis

Management should include the following:

• hospitalization
  
• urine and blood cultures
  
• laboratory studies of complete blood cell count, electrolytes, creatinine, and liver function
  
• monitoring of vital signs and urine output
  
• intravenous (IV) crystalloid fluid to maintain urine output
  
• IV antibiotics (FIGURE 3).
  

If symptoms persist after 48 hours despite adequate IV antibiotic therapy, consider other causes such as resistant organisms, nephrolithiasis, perinephric abscesses, renal obstruction, or other infections.

Consider imaging by renal ultrasound to assess the presence of nephrolithiasis, perinephric abscess, or obstruction.

I recommend inpatient treatment for pregnant women with acute pyelonephritis at this time, until further studies are available.

To evaluate outpatient treatment, Millar and colleagues randomized 120 women under 24 weeks’ gestation either to inpatient IV cefazolin until 48 hours afebrile or to outpatient ceftriaxone intramuscularly. (Both treatment arms completed a course of oral cephalexin.) There were no differences in therapeutic response or birth outcomes, but 6 patients in the outpatient arm required hospitalization for IV therapy and 1 woman developed sepsis.²⁶

The same researchers studied 92 patients of more than 24 weeks’ gestation who received 2 doses of ceftriaxone intramuscularly, then were randomized to either continued inpatient therapy until 48 hours afebrile or discharge with reevaluation as an outpatient in 48 to 72 hours. Again, there were no differences in therapeutic response or birth outcomes; however, almost two-thirds of patients were excluded from the study as they did not meet criteria for outpatient management, due to sepsis, preterm labor, or concurrent medical conditions.²⁷

Adequate antibiotic coverage is crucial

To ensure adequate antibiotic coverage when treating UTI, it is important to understand which organisms cause these infections in pregnancy.

E. coli causes 75% to 90% of UTIs in nonpregnant women.²⁸Staphylococcus saprophyticus causes 10% to 15% of UTIs in nonpregnant women, but less in pregnant women.

Group B Streptococcus (GBS)—another gram-positive organism—has important implications for pregnant women: Intrapartum prophylaxis is important, to prevent neonatal GBS disease.²⁹

Klebsiella, Enterobacter, Proteus, and Enterococcus species²⁸ infrequently cause UTI in pregnancy.

What about prophylaxis and follow-up cultures?

 

Expect recurrence

One third of pregnant women diagnosed with UTI will have recurrence.¹ Recurrence is either relapse (same strain, within 2 weeks of completing initial treatment for the original infection) or reinfection (different strain or same strain after more than 2 weeks).

2 UTIs or pyelonephritis warrant suppressive therapy

I recommend suppressive therapy if a pregnant woman is diagnosed with 2 lower urinary tract infections or acute pyelonephritis (TABLE).

Nitrofurantoin is the preferred agent, as it has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

Start only after eradication of the acute infection, as evidenced by a negative test-of-cure urine culture at least 1 to 2 weeks after treatment is discontinued.²³

Monthly urine cultures until delivery

I recommend monthly follow-up urine cultures until delivery.

Periodic follow-up screening is often recommended, but opinions differ on which test to use or how often to screen.

THE CASE: DIAGNOSIS, TREATMENT, FOLLOW-UP, AND OUTCOME

The patient with upper urinary tract symptoms had a white blood cell count of 15, a urine dipstick positive for leukocyte esterase and nitrites, and a urine sediment analysis indicating pyuria.

She was diagnosed with acute pyelonephritis and started on ampicillin and gentamicin intravenously. Her urine culture drawn upon admission grew >100,000 CFU/mL of sulfonamide-resistant E. coli.

Within 48 hours, she showed clinical improvement and was discharged home with a 10-day course of nitrofurantoin. One week after completing treatment, her test-of-cure urine culture was negative and she was started on nitrofurantoin 50 mg every night at bedtime.

For the rest of pregnancy, she underwent monthly screening urine cultures, which remained negative. She had an uncomplicated delivery at 38 weeks of gestation.

TABLE

Suppressive therapy to prevent UTI recurrence

Suppressive therapy is recommended for any pregnant woman with:
2 lower urinary tract infections or acute pyelonephritis.
Do not initiate suppressive therapy until a negative test-of-cure urine culture confirms eradication of the acute infection.
ANTIBIOTIC	DOSE (ORAL)
Nitrofurantoin monohydrate macrocrystals	50 mg at bedtime
or
Cephalexin	250 mg at bedtime

Dr. Chen reports support from a Women’s Reproductive Health Research Career Development Center grant from the National Institutes of Child Health and Human Development.

THE CASE: SIGNS AND SYMPTOMS

A 29–year–old nullipara at 18 weeks’ gestation complains of fevers and back pain. She had a diagnosis of urinary tract infection with sulfonamide-resistant Escherichia coli at 9 weeks of gestation, which was treated with nitrofurantoin, 100 mg by mouth twice a day for 7 days. A test of cure by urine culture was negative.

Now her temperature is 101°F and she has right costovertebral angle tenderness.

How should you proceed?

Anatomy is destiny, in the case of susceptibility to urinary tract infection (UTI). The female urethra is only 3 cm to 4 cm long, and its proximity to the vagina, anus, and rectum facilitates colonization of normal gastrointestinal flora in the bladder.¹

Sexual activity also facilitates migration of normal gastrointestinal flora to the female urethra.²

Anatomical features of pregnancy exacerbate the female predisposition to urinary tract infection. In pregnancy, hormonal and mechanical changes that occur in the urinary tract lead to urinary stasis and ureterovesical reflux—setting the stage for urinary tract infection (FIGURE 1).

Who should be screened?

All pregnant women should be screened for UTI early in pregnancy, according to the American College of Obstetricians and Gynecologists.³

I recommend a urine culture screening for all pregnant women at their first prenatal visit.

Screen often if she has risk factors

I recommend frequent screening (at least every trimester) by urine culture, in pregnant women with any of these risk factors:

• diabetes mellitus, including gestational diabetes⁴;
  
• urologic abnormalities—specifically, neurogenic bladder;
  
• prepregnancy (for example, 2 to 3 infections per year) and antepartum history of UTI prior to initiation of prenatal care⁵;
  
• sickle cell hemoglobinopathy.⁵
  

Which test is best?

The gold standard for detecting bacteria in urine is by culture.

Which threshold to use?

The standard definition of a positive urine culture from a clean-catch, midstream, voided specimen is ≥100,000 colony forming units (CFU) per mL of a single organism. However, in symptomatic patients, the test’s sensitivity is increased by lowering the cut-off to 100 CFU/mL of a single organism.⁶ In women with urinary symptoms, only 50% of patients had 100,000 CFU/mL by urine culture collected from clean-catch, midstream, voided specimens, though all of them had positive cultures from suprapubic taps.

The clean-catch, midstream, voided specimen is the specimen of choice for practical purposes, since it is noninvasive and easily obtained in the office setting.

For the record: The presence of any organism represents UTI in specimens obtained via suprapubic aspiration of the bladder; 100 CFU/mL of a single organism is positive for specimens obtained by urethral catheterization.

I recommend that, when obtaining urine cultures via clean-catch, midstream, voided specimens:

• for asymptomatic patients, use ≥100,000 CFU/mL of a single organism.
  
• in symptomatic patients, use ≥100 CFU/mL of a single organism.
  

What about rapid tests?

Urinary sediment analysis and urine dipstick testing offer speed and low cost, but with lower accuracy than urine cultures, which require 24 to 48 hours for results and cost more.

Urinary sediment analysis can diagnose pyuria, defined as a clean-catch, midstream, voided specimen, which is spun and which has >10 leukocytes per high-power field.

Pyuria can occur without infection due to:

• previous treatment with antibiotics,
  
• contamination of urine sample by sterilizing solution,
  
• contamination of urine sample with vaginal leukocytes,
  
• chronic interstitial nephritis (such as analgesic abuse),
  
• uroepithelial tumor, and
  
• nephrolithiasis.
  

Pyuria on urinalysis has low sensitivity (25%) but high specificity (99%).

Bacteria visualized on microscopic examination is more sensitive (75%) but less specific (60%).⁷

Urinary dipstick testing—fast, convenient, and low in cost—is considered positive if it identifies either leukocyte esterase or nitrite. Positive leukocyte esterase signifies pyuria. Positive nitrite indicates the presence of enteric organisms that convert urinary nitrate to nitrite.

With either finding, dipstick sensitivity is only 50%, although specificity is 97%.⁷

I recommend:

• If a symptomatic patient’s rapid test is positive, obtain a urine culture, empirically treat for UTI, and then use urine culture results to decide whether to continue treatment.
  
• If an asymptomatic patient’s rapid test is positive, obtain a urine culture and treat only if the culture is positive.
  

What urinary tract disorders occur in pregnancy?

First, determine if the patient has urinary tract symptoms and, if so, whether the symptoms are typical of upper or lower urinary tract infections.

 

Lower urinary tract symptoms:

• dysuria
  
• frequency
  
• urgency
  
• suprapubic pain
  
• hematuria in the absence of fever and systemic symptoms
  

Frequency, of course, is difficult to ascertain in pregnancy, since most women experience this complication secondary to increased urinary output due to higher fluid intake, increased plasma volume expansion, and increases in renal blood flow and glomerular filtration rate.

Upper urinary tract symptoms:

• fever
  
• chills
  
• flank pain
  
• nausea and vomiting
  
• The patient may or may not have the symptoms of lower urinary tract infection, as well.
  

Positive culture and no symptoms

This profile is typical of asymptomatic bacteriuria, a lower urinary tract infection that occurs in 2% to 7% of pregnancies.¹

Positive culture with symptoms

This profile probably reflects either:

• Acute cystitis, a lower urinary tract infection affecting 1% to 2% of pregnancies,⁸ or
  
• Acute pyelonephritis, an upper urinary tract infection affecting 2% of pregnancies.⁹
  

What are the consequences of UTI in pregnancy?

Maternal complications

Asymptomatic bacteriuria does not plague the patient with bothersome effects, but if left untreated, asymptomatic bacteriuria will progress to symptomatic UTI: 25% will develop acute pyelonephritis, compared to 3% to 4% of treated patients¹⁰; 20% of women with severe pyelonephritis develop serious complications,¹¹ including:

• sepsis and septic shock,
  
• hemolysis and thrombocytopenia,¹²
  
• acute respiratory distress syndrome,¹³
  
• renal insufficiency.¹⁴
  

Adverse fetal outcomes

Untreated asymptomatic bacteriuria is associated with preterm delivery and low birthweight.^15-17

Acute pyelonephritis is linked to preterm birth.^18,19 Kaul et al,²⁰ in an experimental model of pyelonephritis in mice, confirmed that E. coli plays an important role in the pathogenesis of preterm delivery and low birthweight.

What is the best treatment regimen?

Data are insufficient to recommend any specific regimen.^21,22 The following strategies are based on evaluation of review articles.^3,23

Asymptomatic bacteriuria and acute cystitis

Nitrofurantoin monohydrate macrocrystals is my first-line treatment. Nitrofurantoin has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

If nitrofurantoin is not effective, I change antibiotics based on urine culture antibiotic sensitivity profiles (FIGURE 2).

Keep in mind the current resistance of E. coli to antibiotics: ampicillin, 28% to 39%; trimethoprim-sulfamethoxazole, 31%; and first-generation cephalosporins, 9% to 19%.²⁴

Single-dose treatment for pregnant women with asymptomatic bacteriuria has been evaluated, given its lower cost and better compliance. However, evidence is insufficient to determine whether single-dose or longer-duration regimens are more effective.²⁵

I recommend longer-duration dosages for now, until a large randomized controlled trial can derive conclusive data.

Remember that treatment success is not contingent upon duration of therapy—just be sure that the test-of-cure urine culture is negative 1 to 2 weeks after treatment is completed.

Acute pyelonephritis

Management should include the following:

• hospitalization
  
• urine and blood cultures
  
• laboratory studies of complete blood cell count, electrolytes, creatinine, and liver function
  
• monitoring of vital signs and urine output
  
• intravenous (IV) crystalloid fluid to maintain urine output
  
• IV antibiotics (FIGURE 3).
  

If symptoms persist after 48 hours despite adequate IV antibiotic therapy, consider other causes such as resistant organisms, nephrolithiasis, perinephric abscesses, renal obstruction, or other infections.

Consider imaging by renal ultrasound to assess the presence of nephrolithiasis, perinephric abscess, or obstruction.

I recommend inpatient treatment for pregnant women with acute pyelonephritis at this time, until further studies are available.

To evaluate outpatient treatment, Millar and colleagues randomized 120 women under 24 weeks’ gestation either to inpatient IV cefazolin until 48 hours afebrile or to outpatient ceftriaxone intramuscularly. (Both treatment arms completed a course of oral cephalexin.) There were no differences in therapeutic response or birth outcomes, but 6 patients in the outpatient arm required hospitalization for IV therapy and 1 woman developed sepsis.²⁶

The same researchers studied 92 patients of more than 24 weeks’ gestation who received 2 doses of ceftriaxone intramuscularly, then were randomized to either continued inpatient therapy until 48 hours afebrile or discharge with reevaluation as an outpatient in 48 to 72 hours. Again, there were no differences in therapeutic response or birth outcomes; however, almost two-thirds of patients were excluded from the study as they did not meet criteria for outpatient management, due to sepsis, preterm labor, or concurrent medical conditions.²⁷

Adequate antibiotic coverage is crucial

To ensure adequate antibiotic coverage when treating UTI, it is important to understand which organisms cause these infections in pregnancy.

E. coli causes 75% to 90% of UTIs in nonpregnant women.²⁸Staphylococcus saprophyticus causes 10% to 15% of UTIs in nonpregnant women, but less in pregnant women.

Group B Streptococcus (GBS)—another gram-positive organism—has important implications for pregnant women: Intrapartum prophylaxis is important, to prevent neonatal GBS disease.²⁹

Klebsiella, Enterobacter, Proteus, and Enterococcus species²⁸ infrequently cause UTI in pregnancy.

What about prophylaxis and follow-up cultures?

 

Expect recurrence

One third of pregnant women diagnosed with UTI will have recurrence.¹ Recurrence is either relapse (same strain, within 2 weeks of completing initial treatment for the original infection) or reinfection (different strain or same strain after more than 2 weeks).

2 UTIs or pyelonephritis warrant suppressive therapy

I recommend suppressive therapy if a pregnant woman is diagnosed with 2 lower urinary tract infections or acute pyelonephritis (TABLE).

Nitrofurantoin is the preferred agent, as it has high concentrations in the urinary tract but induces minimal resistance in gram-negative organisms.

Start only after eradication of the acute infection, as evidenced by a negative test-of-cure urine culture at least 1 to 2 weeks after treatment is discontinued.²³

Monthly urine cultures until delivery

I recommend monthly follow-up urine cultures until delivery.

Periodic follow-up screening is often recommended, but opinions differ on which test to use or how often to screen.

THE CASE: DIAGNOSIS, TREATMENT, FOLLOW-UP, AND OUTCOME

The patient with upper urinary tract symptoms had a white blood cell count of 15, a urine dipstick positive for leukocyte esterase and nitrites, and a urine sediment analysis indicating pyuria.

She was diagnosed with acute pyelonephritis and started on ampicillin and gentamicin intravenously. Her urine culture drawn upon admission grew >100,000 CFU/mL of sulfonamide-resistant E. coli.

Within 48 hours, she showed clinical improvement and was discharged home with a 10-day course of nitrofurantoin. One week after completing treatment, her test-of-cure urine culture was negative and she was started on nitrofurantoin 50 mg every night at bedtime.

For the rest of pregnancy, she underwent monthly screening urine cultures, which remained negative. She had an uncomplicated delivery at 38 weeks of gestation.

TABLE

Suppressive therapy to prevent UTI recurrence

Suppressive therapy is recommended for any pregnant woman with:
2 lower urinary tract infections or acute pyelonephritis.
Do not initiate suppressive therapy until a negative test-of-cure urine culture confirms eradication of the acute infection.
ANTIBIOTIC	DOSE (ORAL)
Nitrofurantoin monohydrate macrocrystals	50 mg at bedtime
or
Cephalexin	250 mg at bedtime

Dr. Chen reports support from a Women’s Reproductive Health Research Career Development Center grant from the National Institutes of Child Health and Human Development.

Although ectopic pregnancy remains a leading cause of life-threatening first-trimester morbidity, accounting for about 9% of maternal deaths annually,¹ we now are able to diagnose and treat most cases well before rupture occurs—in some cases, as early as 5 weeks’ gestation. As a result, medical therapy with systemic methotrexate has become the first-line treatment, with surgery reserved for hemorrhage, medical failures, neglected cases, and circumstances in which medical therapy is contraindicated.

Early diagnosis not only makes medical therapy possible, it also is cheaper, since it avoids rupture, blood loss, and surgery; preserves fertility; and minimizes lost productivity. This is important because ectopic pregnancy is an expensive condition, with an annual health-care bill exceeding $1 billion.²

Despite this progress, serious challenges remain. Medical management is not for everyone. Success is inversely related to initial serum human chorionic gonadotropin (hCG) levels³ and diminishes substantially when embryonic cardiac activity is observed during ultrasound imaging.⁴

In addition, because medical therapy has made outpatient treatment the norm in most cases, it has become virtually impossible to chart the prevalence of ectopic pregnancy. In past years, when hospital records were used, ectopic pregnancy rates were increasing relentlessly, from 4.5 per 1,000 pregnancies in 1970 to 16.8 in 1989 and 19.7 (108,000 cases) in 1992.^1,5

Reasons for increasing rates

Today the prevalence of ectopic pregnancy is probably still rising, for several reasons:

• a greater incidence of risk factors such as sexually transmitted and tubal disease,⁶
• improved diagnostic methods, and
• the use of assisted reproductive technology (ART) to treat infertility (roughly 2% of ART pregnancies are ectopic).⁷

This article describes a 5-step approach to diagnosis and medical management with multiple-dose methotrexate, as well as fine points of treatment and basic surgical technique. It includes a protocol for multiple-dose methotrexate, a table summarizing treatment outcomes, and several case histories.

Likelihood of ectopic pregnancy

If 100 women present with a positive pregnancy test and pain and bleeding, approximately 60 will have a normal pregnancy, 30 are experiencing spontaneous abortion, and 9 have an ectopic pregnancy.⁸

STEP 1Assess risk factors and symptoms

The first step in early diagnosis is being vigilant for risk factors and symptoms associated with ectopic pregnancy, most of which are well known⁹:

Tubal disease carries a 3.5-fold common adjusted odds ratio (OR) for ectopic pregnancy. In addition, women with a previous ectopic pregnancy are 6 to 8 times more likely to experience another, while a history of tubal surgery raises that likelihood to 21. A history of pelvic infection, including gonorrhea, serologically confirmed chlamydia, and pelvic inflammatory disease, increases the risk of ectopic pregnancy 2 to 4 times.⁹

Contraception. Intrauterine devices (IUDs) are associated with an increased OR of 6.4.¹⁰ This does not mean that IUDs cause ectopic pregnancy. Rather, when a woman with an IUD becomes pregnant, an ectopic gestation should be high on the list of possibilities. A similar relationship exists between ectopic pregnancy and tubal ligation, which carries an OR of 9.3.⁹ Oral contraceptives are associated with a reduced risk of ectopic pregnancy unless they are used as emergency contraception (ie, after fertilization), in which case they are associated with an increased risk (TABLE 1).

Diethylstilbestrol exposure in utero alters fallopian tube morphology and can lead to absent or minimal fimbrial tissue, a small tubal os, and decreased length and caliber of the tube.¹¹ Abnormal tubal anatomy caused by this exposure multiplies the risk of ectopic pregnancy by a factor of 5.⁹

In vitro fertilization. When blocked tubes are treated, the embryos can migrate retrograde into the oviduct, implant, and eventually rupture. The OR for ectopic pregnancy with assisted reproduction is 4.0.⁸

 

Cigarette smoking increases the likelihood of ectopic pregnancy 2.5 times,¹² probably by affecting ciliary action within the fallopian tubes.

Salpingitis isthmic nodosa is anatomic thickening of the proximal portion of the fallopian tubes with multiple lumen diverticula. It increases the risk of ectopic pregnancy 1.5 times, compared with age- and race-matched controls.¹³

Don’t depend solely on risk factors. Many ectopic pregnancies present without them.

Symptoms. Many ectopic pregnancies never produce symptoms; rather, they resolve spontaneously or are timely diagnosed and treated medically. Risk factors should therefore be examined in any woman in early pregnancy and investigated further if ectopic pregnancy is likely.

When symptoms do occur, they usually involve 1 or all of the classic triad: amenorrhea, irregular bleeding, and lower abdominal pain. In addition, syncope, shock, and pain radiating to the patient’s shoulder can result from hemoperitoneum.

TABLE 1

High, moderate, and low levels of risk factors for ectopic pregnancy

RISK FACTOR	ODDS RATIO*
High risk
Tubal surgery	21.0
Tubal ligation	9.3
Previous ectopic pregnancy	8.3
In utero exposure to diethylstilbestrol	5.6
Use of intrauterine device	4.2–45.0
Documented tubal pathology	3.8–21.0
Assisted reproduction	4.0
Emergency contraception	High
Moderate risk
Infertility	2.5–21.0
Previous genital infections	2.5–3.7
Multiple sexual partners	2.1
Salpingitis (isthmic)	1.5
Slight risk
Previous pelvic, abdominal surgery	0.9–3.8
Cigarette smoking	2.3–2.5
Vaginal douching	1.1–3.1
Early age at first intercourse (<18 years)	1.6
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).
* Single values = common odds ratio from homogeneous studies; point estimates = range of values from heterogeneous studies

STEP 2Document the pregnancy and measure ß-hCG

Once you identify the high-risk patient, or a woman comes in complaining of pain and spotting or bleeding, run a pregnancy test to confirm that she is pregnant and, if it is positive, obtain a quantitative ß-hCG.

ß-hCG levels are normally measured using enzyme-linked immunosorbent assays (ELISA), which detect ß-hCG in urine and serum at levels as low as 20 mIU/mL and 10 mIU/mL, respectively.¹⁴ ß-hCG is produced by trophoblastic cells in normal pregnancy, and approximately doubles every 2 days when titers are below 10,000 mIU/mL¹⁵—although in some normal pregnancies, ß-hCG may increase as slowly as 53% or as rapidly as 230% over 2 days.¹⁶ Eighty-five percent of abnormal pregnancies—whether intrauterine or ectopic—have impaired ß-hCG production with prolonged doubling time. Thus, in failing pregnancies, ß-hCG levels will plateau or fail to rise normally.

A single ß-hCG level fails to predict the risk of rupture, since ectopic pregnancies can rupture at ß-hCG levels as low as 10 mIU/mL or far exceeding 10,000 mIU/mL, or at any level in between.

STEP 3Obtain an ultrasound scan

Transvaginal ultrasound reliably detects normal intrauterine gestations when ß-hCG passes somewhere between 1,000 mIU/mL and 2,000 mIU/mL (First International Reference Preparation), depending on the expertise of the ultrasonographer and the particular equipment used.^8,17 This is known as the “discriminatory zone.” ß-hCG levels reach this zone as early as 1 week after missed menses.¹⁸

The discriminatory zone is not the lowest ß-hCG concentration at which an intrauterine pregnancy can be visualized via ultrasound. Rather, it is the value at which any intrauterine pregnancy will be apparent. At that value, the absence of an intrauterine pregnancy confirms—by negative conclusion—that the patient has a nonviable gestation.

When intrauterine pregnancy is visualized. The diagnosis is definitive and the woman’s symptoms can be explained as “threatened abortion.” No further investigation is necessary aside from routine prenatal care if the pregnancy continues.

When an extrauterine gestation is observed, such as a gestational sac with a detectable fetal heart rate, ectopic pregnancy can be diagnosed with 100% specificity but low sensitivity (15% to 20%). A complex adnexal mass without an intrauterine pregnancy improves sensitivity from 21% to 84% at the expense of lower specificity (93% to 99.5%).¹⁹

Even when an adnexal mass is visualized, cardiac activity is not usually present. If cardiac activity is apparent, proceed to surgery, since methotrexate usually will not resolve these gestations.

Be aware that some adnexal masses suspicious for ectopic pregnancy may turn out to be other entities, such as a corpus luteum, hydrosalpinx, ovarian neoplasm, or endometrioma. Unless a fetal heart rate is detected by ultrasound, the diagnosis is uncertain and curettage is needed to establish a definitive diagnosis.

No intrauterine pregnancy, no extrauterine mass. Despite the high resolution of transvaginal ultrasound, many patients with ectopic pregnancy have no apparent adnexal mass,²⁰ particularly when diagnosis is early. In these cases, proceed to curettage (step 4).

Don’t interpret ultrasound findings in a vacuum

This is especially unwise when ß-hCG levels are low—even when the ultrasound report points to intrauterine pregnancy. At ß-hCG levels below 1,500 mIU/mL, the sensitivity of ultrasound in diagnosing intrauterine pregnancy drops from 98% to 33% and predictive value is substantially lower. Interpret ultrasound and ß-hCG levels together for greater accuracy.

 

How size influences management

Ultrasound can detect ectopic pregnancies as small as 2 cm. In general, an ectopic sac size larger than 4 cm should be treated surgically.

STEP 4Perform uterine curettage

If ultrasound imaging is inconclusive and ß-hCG levels are plateauing or rising subnormally, perform uterine curettage. If ß-hCG levels decrease 15% or more 8 to 12 hours after the procedure, a complete abortion can be strongly suspected.²¹ If ß-hCG levels plateau or rise, the trophoblasts were not removed by curettage, and ectopic pregnancy is diagnosed.²¹

Keep in mind these important points:

• Without uterine curettage, roughly 40% of ectopic pregnancy diagnoses are incorrect.²²
• Because curettage will result in termination of pregnancy, it is vital that it be limited to cases involving abnormal ß-hCG levels, not normally rising values.

STEP 5Administer methotrexate

Medical management is indicated when the following circumstances are present:

• No viable intrauterine pregnancy is present.
• No rupture has occurred.
• Any adnexal mass is 4 cm in size or smaller.
• ß-hCG levels are below 10,000 mIU/mL.

If there is a positive fetal heart beat, surgery is preferred.

Firm diagnosis of ectopic pregnancy is essential prior to methotrexate administration. If the drug is given to a woman carrying a viable pregnancy, it may result in loss of the pregnancy or methotrexate embryopathy.²³

How methotrexate resolves ectopic pregnancy

Systemic methotrexate has been used successfully in the treatment of ectopic pregnancy for more than 20 years. A folic acid antagonist, it inhibits de novo synthesis of purines and pyrimidines. It thus interferes with DNA synthesis and cell multiplication.²⁴ Actively proliferating trophoblasts are particularly vulnerable.²⁵

When methotrexate is administered to normally pregnant women, it blunts the normal ß-hCG increment over the next 7 days. Circulating progesterone and 17-a-hydroxyprogesterone concentrations also decline, and abortion occurs.²⁶

Methotrexate directly impairs trophoblastic production of hCG; the decrement in corpus luteum progesterone is a secondary event.

Side effects include abdominal distress, chills and fever, dizziness, immunosuppression, leukopenia, malaise, nausea, ulcerative stomatitis, photosensitivity, and undue fatigue.

Breastfeeding is an absolute contraindication to methotrexate, while relative contraindications include abnormal liver function tests, blood dyscrasias, excessive alcohol consumption, HIV/AIDS, psoriasis, ongoing radiotherapy, rheumatoid arthritis, and significant pulmonary disease.

Multiple-dose methotrexate is superior

In a recent meta-analysis comparing single-and multiple-dose regimens of methotrexate, Barnhart and colleagues⁴ found the latter to be more effective, although the single dose was more commonly given. The single-dose regimen was associated with a significantly greater chance of failure in both crude (OR 1.71; 1.04, 2.82) and adjusted (OR 4.74; 1.77, 12.62) analyses. Consequently, we advocate the multiple-dose regimen detailed in TABLE 2: 1 mg of methotrexate per kilogram of body weight on day 1, alternating with 0.1 mg of leucovorin per kilogram on succeeding days. Continue this regimen until ß-hCG levels decline in 2 consecutive daily titers, or 4 doses of methotrexate are given, whichever comes first.

If treatment is unsuccessful after 4 doses, additional methotrexate is unlikely to be effective and will involve significant additional cost and morbidity. If ß-hCG levels plateau or continue to rise, surgery is indicated.

Although multiple-dose therapy is more effective than a single dose, the optimal number of doses probably falls somewhere between 1 and 4.⁴ A 2-dose protocol currently under investigation may provide an optimal compromise.

Artificially high efficacy rates with “single-dose” therapy. Unusually high success rates in initial studies of the single-dose regimen may have been due to the inclusion of spontaneously aborting intrauterine pregnancies.²⁷ Although 6 subsequent studies—1 cohort and 5 case-control studies involving 304 patients—found an overall success rate (no surgical intervention) of 87.2%, 11.5% of participants required more than 1 dose.⁹

Still the standard. Despite its lower efficacy rates, single-dose methotrexate remains the “standard” in the United States, as recommended in an American College of Obstetricians and Gynecologists practice bulletin.²⁸ The usual intramuscular dose is 50 mg per square meter of body surface area, with ß-hCG titers on days 4 and 7 and an additional dose if ß-hCG levels fail to resolve.

What the evidence shows

Multidose methotrexate. Twelve studies measured the success of multiple-dose systemic methotrexate (TABLE 3); they included 1 randomized, controlled trial, 1 cohort study, and 10 case series. Between 1982 and 1997, 325 cases were treated with multiple-dose methotrexate. Of these, 93.8% had successful resolution with no subsequent therapy, and 78.9% of the 161 women tested had patent oviducts. In addition, of 95 women hoping to conceive, 57.9% had a subsequent intrauterine pregnancy and 7.4% developed a recurrent ectopic pregnancy. These rates compared favorably with numerous laparoscopic surgery series published during the same years.⁹

 

Methotrexate versus laparoscopic salpingostomy. In 1 randomized clinical comparison,²⁹ 100 women with laparoscopy-confirmed ectopic pregnancy were randomized to methotrexate or laparoscopic salpingostomy. Of the 51 patients treated medically, 7 (14%) required surgical intervention for active bleeding and/or tubal rupture. An additional course of methotrexate was required in 2 patients (4%) for persistent trophoblasts.

Of the 49 patients in the salpingostomy group, 4 women (8%) failed therapy and required salpingectomies, and 10 patients (20%) were treated with methotrexate for persistent trophoblasts.

Tubal patency was present in 23 of 42 women (55%) in the methotrexate group, compared with 23 of 39 (59%) in the salpingostomy group.

Overall, this randomized study²⁹ and previous meta-analysis demonstrate that systemic multiple-dose methotrexate is comparable in efficacy to laparoscopic salpingostomy.

TABLE 2

Multiple-dose methotrexate protocol

Discontinue treatment when there is a decline in 2 consecutive ß-hCG titers or after 4 doses, whichever comes first.

DAY	INTERVENTION	DOSE (MG)
1	Baseline studies ß-hCG titer, CBC, and platelets Methotrexate	1.0
2	Leucovori	0.1
3	Methotrexate	1.0
4	Leucovorin ß-hCG titer	0.1
5	Methotrexate ß-hCG titer	1.0
6	Leucovorin ß-hCG titer	0.1
7	Methotrexate ß-hCG titer	1.0
8	Leucovorin ß-hCG titer CBC and platelets Renal and liver function tests	0.1
Weekly	ß-hCG titer until negative

TABLE 3

Treatment outcomes for ectopic pregnancy

METHOD	NUMBER OF STUDIES	NUMBER OF PATIENTS	NUMBER WITH SUCCESSFUL RESOLUTION	TUBAL PATENCY RATE	SUBSEQUENT FERTILITY RATE
					INTRAUTERINE PREGNANCY	ECTOPIC PREGNANCY
Conservative laparoscopic surgery	32	1,626	1,516 (93%)	170/223 (76%)	366/647 (57%)	87/647 (13%)
Variable-dose methotrexate	12	338	314 (93%)	136/182 (75%)	55/95 (58%)	7/95 (7%)
Single-dose methotrexate	7	393	340 (87%)	61/75 (81%)	39/64 (61%)	5/64 (8%)
Direct-injection methotrexate	21	660	502 (76%)	130/162 (80%)	87/152 (57%)	9/152 (6%)
Expectant management	14	628	425 (68%)	60/79 (76%)	12/14 (86%)	1/14 (7%)
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).

Fine points of treatment

During methotrexate therapy, examine the patient only once to avoid triggering a rupture, and counsel her to avoid intercourse for the same reason.³⁰ Do not perform repeat vaginal ultrasound examination. Also inform her that transient pain (“separation pain”) from tubal abortion frequently occurs 3 to 7 days after the start of therapy, lasts 4 to 12 hours, and then resolves.³¹ This is perhaps the most difficult aspect of methotrexate therapy, as it is not always easy to differentiate the pain of tubal abortion from the pain of rupture.

If ß-hCG titers continue to rise rapidly between methotrexate doses, rupture is more likely and surgery should proceed.³²

Overall, isthmic ectopic pregnancies (12.3% of ectopics) appear to be at a particularly high risk for rupture and comprise nearly half of methotrexate failures.³² Unfortunately, there is no way to identify isthmic pregnancies without surgery.

Avoid NSAIDs and GI-“unfriendly” foods. Counsel the patient to avoid nonsteroidal anti-inflammatory drugs (NSAIDs) because they may impair natural hemostasis. Gasforming foods such as leeks, corn, and cabbage can cause distension, which may be mistaken for rupture.

Also instruct the patient to avoid folic acid, which impairs the efficacy of methotrexate.

Ultrasound surveillance is unnecessary. If an adnexal mass was identified at initial imaging, there is no need to view it again, since these masses tend to enlarge and form hematomas and can cause undue anxiety in both physician and patient. In properly selected patients, multidose methotrexate with monitoring of ß-hCG levels should suffice.

When surgery is indicated

Surgical intervention is necessary when pain is severe, persists beyond 12 hours, and is associated with orthostatic hypotension, falling hematocrit, or persistently elevated ß-hCG levels after methotrexate therapy.

Laparoscopy is the preferred approach. Advantages include less blood loss and analgesia,³³ shortened postoperative recovery, and lower costs.

Technique. For unruptured ampullary ectopic pregnancy, salpingostomy is preferred. Make a linear incision over the bulging antimesenteric border of the fallopian tube using electrocautery, scissors, or laser. Remove the products of conception using forceps or suction, and leave the incision to heal by secondary intention.

For isthmic pregnancies, use segmental excision followed by delayed microsurgical anastomosis.³⁴ The isthmic tubal lumen is narrower and the muscularis thicker than in the ampulla. Thus, the isthmus is predisposed to greater damage after salpingostomy and greater rates of proximal obstruction.³³

An increased risk of persistent ectopic pregnancy has been a criticism of salpingostomy. However, when 1 dose of systemic methotrexate is combined with salpingostomy, the risk of persistent pregnancy is virtually eliminated.³⁵

When rupture occurs, salpingectomy is the first choice for treatment, as it arrests hemorrhage and shortens the procedure. Either laparoscopy or laparotomy is appropriate.

The authors report no financial relationships relevant to this article.

Although ectopic pregnancy remains a leading cause of life-threatening first-trimester morbidity, accounting for about 9% of maternal deaths annually,¹ we now are able to diagnose and treat most cases well before rupture occurs—in some cases, as early as 5 weeks’ gestation. As a result, medical therapy with systemic methotrexate has become the first-line treatment, with surgery reserved for hemorrhage, medical failures, neglected cases, and circumstances in which medical therapy is contraindicated.

Early diagnosis not only makes medical therapy possible, it also is cheaper, since it avoids rupture, blood loss, and surgery; preserves fertility; and minimizes lost productivity. This is important because ectopic pregnancy is an expensive condition, with an annual health-care bill exceeding $1 billion.²

Despite this progress, serious challenges remain. Medical management is not for everyone. Success is inversely related to initial serum human chorionic gonadotropin (hCG) levels³ and diminishes substantially when embryonic cardiac activity is observed during ultrasound imaging.⁴

In addition, because medical therapy has made outpatient treatment the norm in most cases, it has become virtually impossible to chart the prevalence of ectopic pregnancy. In past years, when hospital records were used, ectopic pregnancy rates were increasing relentlessly, from 4.5 per 1,000 pregnancies in 1970 to 16.8 in 1989 and 19.7 (108,000 cases) in 1992.^1,5

Reasons for increasing rates

Today the prevalence of ectopic pregnancy is probably still rising, for several reasons:

• a greater incidence of risk factors such as sexually transmitted and tubal disease,⁶
• improved diagnostic methods, and
• the use of assisted reproductive technology (ART) to treat infertility (roughly 2% of ART pregnancies are ectopic).⁷

This article describes a 5-step approach to diagnosis and medical management with multiple-dose methotrexate, as well as fine points of treatment and basic surgical technique. It includes a protocol for multiple-dose methotrexate, a table summarizing treatment outcomes, and several case histories.

Likelihood of ectopic pregnancy

If 100 women present with a positive pregnancy test and pain and bleeding, approximately 60 will have a normal pregnancy, 30 are experiencing spontaneous abortion, and 9 have an ectopic pregnancy.⁸

STEP 1Assess risk factors and symptoms

The first step in early diagnosis is being vigilant for risk factors and symptoms associated with ectopic pregnancy, most of which are well known⁹:

Tubal disease carries a 3.5-fold common adjusted odds ratio (OR) for ectopic pregnancy. In addition, women with a previous ectopic pregnancy are 6 to 8 times more likely to experience another, while a history of tubal surgery raises that likelihood to 21. A history of pelvic infection, including gonorrhea, serologically confirmed chlamydia, and pelvic inflammatory disease, increases the risk of ectopic pregnancy 2 to 4 times.⁹

Contraception. Intrauterine devices (IUDs) are associated with an increased OR of 6.4.¹⁰ This does not mean that IUDs cause ectopic pregnancy. Rather, when a woman with an IUD becomes pregnant, an ectopic gestation should be high on the list of possibilities. A similar relationship exists between ectopic pregnancy and tubal ligation, which carries an OR of 9.3.⁹ Oral contraceptives are associated with a reduced risk of ectopic pregnancy unless they are used as emergency contraception (ie, after fertilization), in which case they are associated with an increased risk (TABLE 1).

Diethylstilbestrol exposure in utero alters fallopian tube morphology and can lead to absent or minimal fimbrial tissue, a small tubal os, and decreased length and caliber of the tube.¹¹ Abnormal tubal anatomy caused by this exposure multiplies the risk of ectopic pregnancy by a factor of 5.⁹

In vitro fertilization. When blocked tubes are treated, the embryos can migrate retrograde into the oviduct, implant, and eventually rupture. The OR for ectopic pregnancy with assisted reproduction is 4.0.⁸

 

Cigarette smoking increases the likelihood of ectopic pregnancy 2.5 times,¹² probably by affecting ciliary action within the fallopian tubes.

Salpingitis isthmic nodosa is anatomic thickening of the proximal portion of the fallopian tubes with multiple lumen diverticula. It increases the risk of ectopic pregnancy 1.5 times, compared with age- and race-matched controls.¹³

Don’t depend solely on risk factors. Many ectopic pregnancies present without them.

Symptoms. Many ectopic pregnancies never produce symptoms; rather, they resolve spontaneously or are timely diagnosed and treated medically. Risk factors should therefore be examined in any woman in early pregnancy and investigated further if ectopic pregnancy is likely.

When symptoms do occur, they usually involve 1 or all of the classic triad: amenorrhea, irregular bleeding, and lower abdominal pain. In addition, syncope, shock, and pain radiating to the patient’s shoulder can result from hemoperitoneum.

TABLE 1

High, moderate, and low levels of risk factors for ectopic pregnancy

RISK FACTOR	ODDS RATIO*
High risk
Tubal surgery	21.0
Tubal ligation	9.3
Previous ectopic pregnancy	8.3
In utero exposure to diethylstilbestrol	5.6
Use of intrauterine device	4.2–45.0
Documented tubal pathology	3.8–21.0
Assisted reproduction	4.0
Emergency contraception	High
Moderate risk
Infertility	2.5–21.0
Previous genital infections	2.5–3.7
Multiple sexual partners	2.1
Salpingitis (isthmic)	1.5
Slight risk
Previous pelvic, abdominal surgery	0.9–3.8
Cigarette smoking	2.3–2.5
Vaginal douching	1.1–3.1
Early age at first intercourse (<18 years)	1.6
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).
* Single values = common odds ratio from homogeneous studies; point estimates = range of values from heterogeneous studies

STEP 2Document the pregnancy and measure ß-hCG

Once you identify the high-risk patient, or a woman comes in complaining of pain and spotting or bleeding, run a pregnancy test to confirm that she is pregnant and, if it is positive, obtain a quantitative ß-hCG.

ß-hCG levels are normally measured using enzyme-linked immunosorbent assays (ELISA), which detect ß-hCG in urine and serum at levels as low as 20 mIU/mL and 10 mIU/mL, respectively.¹⁴ ß-hCG is produced by trophoblastic cells in normal pregnancy, and approximately doubles every 2 days when titers are below 10,000 mIU/mL¹⁵—although in some normal pregnancies, ß-hCG may increase as slowly as 53% or as rapidly as 230% over 2 days.¹⁶ Eighty-five percent of abnormal pregnancies—whether intrauterine or ectopic—have impaired ß-hCG production with prolonged doubling time. Thus, in failing pregnancies, ß-hCG levels will plateau or fail to rise normally.

A single ß-hCG level fails to predict the risk of rupture, since ectopic pregnancies can rupture at ß-hCG levels as low as 10 mIU/mL or far exceeding 10,000 mIU/mL, or at any level in between.

STEP 3Obtain an ultrasound scan

Transvaginal ultrasound reliably detects normal intrauterine gestations when ß-hCG passes somewhere between 1,000 mIU/mL and 2,000 mIU/mL (First International Reference Preparation), depending on the expertise of the ultrasonographer and the particular equipment used.^8,17 This is known as the “discriminatory zone.” ß-hCG levels reach this zone as early as 1 week after missed menses.¹⁸

The discriminatory zone is not the lowest ß-hCG concentration at which an intrauterine pregnancy can be visualized via ultrasound. Rather, it is the value at which any intrauterine pregnancy will be apparent. At that value, the absence of an intrauterine pregnancy confirms—by negative conclusion—that the patient has a nonviable gestation.

When intrauterine pregnancy is visualized. The diagnosis is definitive and the woman’s symptoms can be explained as “threatened abortion.” No further investigation is necessary aside from routine prenatal care if the pregnancy continues.

When an extrauterine gestation is observed, such as a gestational sac with a detectable fetal heart rate, ectopic pregnancy can be diagnosed with 100% specificity but low sensitivity (15% to 20%). A complex adnexal mass without an intrauterine pregnancy improves sensitivity from 21% to 84% at the expense of lower specificity (93% to 99.5%).¹⁹

Even when an adnexal mass is visualized, cardiac activity is not usually present. If cardiac activity is apparent, proceed to surgery, since methotrexate usually will not resolve these gestations.

Be aware that some adnexal masses suspicious for ectopic pregnancy may turn out to be other entities, such as a corpus luteum, hydrosalpinx, ovarian neoplasm, or endometrioma. Unless a fetal heart rate is detected by ultrasound, the diagnosis is uncertain and curettage is needed to establish a definitive diagnosis.

No intrauterine pregnancy, no extrauterine mass. Despite the high resolution of transvaginal ultrasound, many patients with ectopic pregnancy have no apparent adnexal mass,²⁰ particularly when diagnosis is early. In these cases, proceed to curettage (step 4).

Don’t interpret ultrasound findings in a vacuum

This is especially unwise when ß-hCG levels are low—even when the ultrasound report points to intrauterine pregnancy. At ß-hCG levels below 1,500 mIU/mL, the sensitivity of ultrasound in diagnosing intrauterine pregnancy drops from 98% to 33% and predictive value is substantially lower. Interpret ultrasound and ß-hCG levels together for greater accuracy.

 

How size influences management

Ultrasound can detect ectopic pregnancies as small as 2 cm. In general, an ectopic sac size larger than 4 cm should be treated surgically.

STEP 4Perform uterine curettage

If ultrasound imaging is inconclusive and ß-hCG levels are plateauing or rising subnormally, perform uterine curettage. If ß-hCG levels decrease 15% or more 8 to 12 hours after the procedure, a complete abortion can be strongly suspected.²¹ If ß-hCG levels plateau or rise, the trophoblasts were not removed by curettage, and ectopic pregnancy is diagnosed.²¹

Keep in mind these important points:

• Without uterine curettage, roughly 40% of ectopic pregnancy diagnoses are incorrect.²²
• Because curettage will result in termination of pregnancy, it is vital that it be limited to cases involving abnormal ß-hCG levels, not normally rising values.

STEP 5Administer methotrexate

Medical management is indicated when the following circumstances are present:

• No viable intrauterine pregnancy is present.
• No rupture has occurred.
• Any adnexal mass is 4 cm in size or smaller.
• ß-hCG levels are below 10,000 mIU/mL.

If there is a positive fetal heart beat, surgery is preferred.

Firm diagnosis of ectopic pregnancy is essential prior to methotrexate administration. If the drug is given to a woman carrying a viable pregnancy, it may result in loss of the pregnancy or methotrexate embryopathy.²³

How methotrexate resolves ectopic pregnancy

Systemic methotrexate has been used successfully in the treatment of ectopic pregnancy for more than 20 years. A folic acid antagonist, it inhibits de novo synthesis of purines and pyrimidines. It thus interferes with DNA synthesis and cell multiplication.²⁴ Actively proliferating trophoblasts are particularly vulnerable.²⁵

When methotrexate is administered to normally pregnant women, it blunts the normal ß-hCG increment over the next 7 days. Circulating progesterone and 17-a-hydroxyprogesterone concentrations also decline, and abortion occurs.²⁶

Methotrexate directly impairs trophoblastic production of hCG; the decrement in corpus luteum progesterone is a secondary event.

Side effects include abdominal distress, chills and fever, dizziness, immunosuppression, leukopenia, malaise, nausea, ulcerative stomatitis, photosensitivity, and undue fatigue.

Breastfeeding is an absolute contraindication to methotrexate, while relative contraindications include abnormal liver function tests, blood dyscrasias, excessive alcohol consumption, HIV/AIDS, psoriasis, ongoing radiotherapy, rheumatoid arthritis, and significant pulmonary disease.

Multiple-dose methotrexate is superior

In a recent meta-analysis comparing single-and multiple-dose regimens of methotrexate, Barnhart and colleagues⁴ found the latter to be more effective, although the single dose was more commonly given. The single-dose regimen was associated with a significantly greater chance of failure in both crude (OR 1.71; 1.04, 2.82) and adjusted (OR 4.74; 1.77, 12.62) analyses. Consequently, we advocate the multiple-dose regimen detailed in TABLE 2: 1 mg of methotrexate per kilogram of body weight on day 1, alternating with 0.1 mg of leucovorin per kilogram on succeeding days. Continue this regimen until ß-hCG levels decline in 2 consecutive daily titers, or 4 doses of methotrexate are given, whichever comes first.

If treatment is unsuccessful after 4 doses, additional methotrexate is unlikely to be effective and will involve significant additional cost and morbidity. If ß-hCG levels plateau or continue to rise, surgery is indicated.

Although multiple-dose therapy is more effective than a single dose, the optimal number of doses probably falls somewhere between 1 and 4.⁴ A 2-dose protocol currently under investigation may provide an optimal compromise.

Artificially high efficacy rates with “single-dose” therapy. Unusually high success rates in initial studies of the single-dose regimen may have been due to the inclusion of spontaneously aborting intrauterine pregnancies.²⁷ Although 6 subsequent studies—1 cohort and 5 case-control studies involving 304 patients—found an overall success rate (no surgical intervention) of 87.2%, 11.5% of participants required more than 1 dose.⁹

Still the standard. Despite its lower efficacy rates, single-dose methotrexate remains the “standard” in the United States, as recommended in an American College of Obstetricians and Gynecologists practice bulletin.²⁸ The usual intramuscular dose is 50 mg per square meter of body surface area, with ß-hCG titers on days 4 and 7 and an additional dose if ß-hCG levels fail to resolve.

What the evidence shows

Multidose methotrexate. Twelve studies measured the success of multiple-dose systemic methotrexate (TABLE 3); they included 1 randomized, controlled trial, 1 cohort study, and 10 case series. Between 1982 and 1997, 325 cases were treated with multiple-dose methotrexate. Of these, 93.8% had successful resolution with no subsequent therapy, and 78.9% of the 161 women tested had patent oviducts. In addition, of 95 women hoping to conceive, 57.9% had a subsequent intrauterine pregnancy and 7.4% developed a recurrent ectopic pregnancy. These rates compared favorably with numerous laparoscopic surgery series published during the same years.⁹

 

Methotrexate versus laparoscopic salpingostomy. In 1 randomized clinical comparison,²⁹ 100 women with laparoscopy-confirmed ectopic pregnancy were randomized to methotrexate or laparoscopic salpingostomy. Of the 51 patients treated medically, 7 (14%) required surgical intervention for active bleeding and/or tubal rupture. An additional course of methotrexate was required in 2 patients (4%) for persistent trophoblasts.

Of the 49 patients in the salpingostomy group, 4 women (8%) failed therapy and required salpingectomies, and 10 patients (20%) were treated with methotrexate for persistent trophoblasts.

Tubal patency was present in 23 of 42 women (55%) in the methotrexate group, compared with 23 of 39 (59%) in the salpingostomy group.

Overall, this randomized study²⁹ and previous meta-analysis demonstrate that systemic multiple-dose methotrexate is comparable in efficacy to laparoscopic salpingostomy.

TABLE 2

Multiple-dose methotrexate protocol

Discontinue treatment when there is a decline in 2 consecutive ß-hCG titers or after 4 doses, whichever comes first.

DAY	INTERVENTION	DOSE (MG)
1	Baseline studies ß-hCG titer, CBC, and platelets Methotrexate	1.0
2	Leucovori	0.1
3	Methotrexate	1.0
4	Leucovorin ß-hCG titer	0.1
5	Methotrexate ß-hCG titer	1.0
6	Leucovorin ß-hCG titer	0.1
7	Methotrexate ß-hCG titer	1.0
8	Leucovorin ß-hCG titer CBC and platelets Renal and liver function tests	0.1
Weekly	ß-hCG titer until negative

TABLE 3

Treatment outcomes for ectopic pregnancy

METHOD	NUMBER OF STUDIES	NUMBER OF PATIENTS	NUMBER WITH SUCCESSFUL RESOLUTION	TUBAL PATENCY RATE	SUBSEQUENT FERTILITY RATE
					INTRAUTERINE PREGNANCY	ECTOPIC PREGNANCY
Conservative laparoscopic surgery	32	1,626	1,516 (93%)	170/223 (76%)	366/647 (57%)	87/647 (13%)
Variable-dose methotrexate	12	338	314 (93%)	136/182 (75%)	55/95 (58%)	7/95 (7%)
Single-dose methotrexate	7	393	340 (87%)	61/75 (81%)	39/64 (61%)	5/64 (8%)
Direct-injection methotrexate	21	660	502 (76%)	130/162 (80%)	87/152 (57%)	9/152 (6%)
Expectant management	14	628	425 (68%)	60/79 (76%)	12/14 (86%)	1/14 (7%)
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).

Fine points of treatment

During methotrexate therapy, examine the patient only once to avoid triggering a rupture, and counsel her to avoid intercourse for the same reason.³⁰ Do not perform repeat vaginal ultrasound examination. Also inform her that transient pain (“separation pain”) from tubal abortion frequently occurs 3 to 7 days after the start of therapy, lasts 4 to 12 hours, and then resolves.³¹ This is perhaps the most difficult aspect of methotrexate therapy, as it is not always easy to differentiate the pain of tubal abortion from the pain of rupture.

If ß-hCG titers continue to rise rapidly between methotrexate doses, rupture is more likely and surgery should proceed.³²

Overall, isthmic ectopic pregnancies (12.3% of ectopics) appear to be at a particularly high risk for rupture and comprise nearly half of methotrexate failures.³² Unfortunately, there is no way to identify isthmic pregnancies without surgery.

Avoid NSAIDs and GI-“unfriendly” foods. Counsel the patient to avoid nonsteroidal anti-inflammatory drugs (NSAIDs) because they may impair natural hemostasis. Gasforming foods such as leeks, corn, and cabbage can cause distension, which may be mistaken for rupture.

Also instruct the patient to avoid folic acid, which impairs the efficacy of methotrexate.

Ultrasound surveillance is unnecessary. If an adnexal mass was identified at initial imaging, there is no need to view it again, since these masses tend to enlarge and form hematomas and can cause undue anxiety in both physician and patient. In properly selected patients, multidose methotrexate with monitoring of ß-hCG levels should suffice.

When surgery is indicated

Surgical intervention is necessary when pain is severe, persists beyond 12 hours, and is associated with orthostatic hypotension, falling hematocrit, or persistently elevated ß-hCG levels after methotrexate therapy.

Laparoscopy is the preferred approach. Advantages include less blood loss and analgesia,³³ shortened postoperative recovery, and lower costs.

Technique. For unruptured ampullary ectopic pregnancy, salpingostomy is preferred. Make a linear incision over the bulging antimesenteric border of the fallopian tube using electrocautery, scissors, or laser. Remove the products of conception using forceps or suction, and leave the incision to heal by secondary intention.

For isthmic pregnancies, use segmental excision followed by delayed microsurgical anastomosis.³⁴ The isthmic tubal lumen is narrower and the muscularis thicker than in the ampulla. Thus, the isthmus is predisposed to greater damage after salpingostomy and greater rates of proximal obstruction.³³

An increased risk of persistent ectopic pregnancy has been a criticism of salpingostomy. However, when 1 dose of systemic methotrexate is combined with salpingostomy, the risk of persistent pregnancy is virtually eliminated.³⁵

When rupture occurs, salpingectomy is the first choice for treatment, as it arrests hemorrhage and shortens the procedure. Either laparoscopy or laparotomy is appropriate.

The authors report no financial relationships relevant to this article.

Although ectopic pregnancy remains a leading cause of life-threatening first-trimester morbidity, accounting for about 9% of maternal deaths annually,¹ we now are able to diagnose and treat most cases well before rupture occurs—in some cases, as early as 5 weeks’ gestation. As a result, medical therapy with systemic methotrexate has become the first-line treatment, with surgery reserved for hemorrhage, medical failures, neglected cases, and circumstances in which medical therapy is contraindicated.

Early diagnosis not only makes medical therapy possible, it also is cheaper, since it avoids rupture, blood loss, and surgery; preserves fertility; and minimizes lost productivity. This is important because ectopic pregnancy is an expensive condition, with an annual health-care bill exceeding $1 billion.²

Despite this progress, serious challenges remain. Medical management is not for everyone. Success is inversely related to initial serum human chorionic gonadotropin (hCG) levels³ and diminishes substantially when embryonic cardiac activity is observed during ultrasound imaging.⁴

In addition, because medical therapy has made outpatient treatment the norm in most cases, it has become virtually impossible to chart the prevalence of ectopic pregnancy. In past years, when hospital records were used, ectopic pregnancy rates were increasing relentlessly, from 4.5 per 1,000 pregnancies in 1970 to 16.8 in 1989 and 19.7 (108,000 cases) in 1992.^1,5

Reasons for increasing rates

Today the prevalence of ectopic pregnancy is probably still rising, for several reasons:

• a greater incidence of risk factors such as sexually transmitted and tubal disease,⁶
• improved diagnostic methods, and
• the use of assisted reproductive technology (ART) to treat infertility (roughly 2% of ART pregnancies are ectopic).⁷

This article describes a 5-step approach to diagnosis and medical management with multiple-dose methotrexate, as well as fine points of treatment and basic surgical technique. It includes a protocol for multiple-dose methotrexate, a table summarizing treatment outcomes, and several case histories.

Likelihood of ectopic pregnancy

If 100 women present with a positive pregnancy test and pain and bleeding, approximately 60 will have a normal pregnancy, 30 are experiencing spontaneous abortion, and 9 have an ectopic pregnancy.⁸

STEP 1Assess risk factors and symptoms

The first step in early diagnosis is being vigilant for risk factors and symptoms associated with ectopic pregnancy, most of which are well known⁹:

Tubal disease carries a 3.5-fold common adjusted odds ratio (OR) for ectopic pregnancy. In addition, women with a previous ectopic pregnancy are 6 to 8 times more likely to experience another, while a history of tubal surgery raises that likelihood to 21. A history of pelvic infection, including gonorrhea, serologically confirmed chlamydia, and pelvic inflammatory disease, increases the risk of ectopic pregnancy 2 to 4 times.⁹

Contraception. Intrauterine devices (IUDs) are associated with an increased OR of 6.4.¹⁰ This does not mean that IUDs cause ectopic pregnancy. Rather, when a woman with an IUD becomes pregnant, an ectopic gestation should be high on the list of possibilities. A similar relationship exists between ectopic pregnancy and tubal ligation, which carries an OR of 9.3.⁹ Oral contraceptives are associated with a reduced risk of ectopic pregnancy unless they are used as emergency contraception (ie, after fertilization), in which case they are associated with an increased risk (TABLE 1).

Diethylstilbestrol exposure in utero alters fallopian tube morphology and can lead to absent or minimal fimbrial tissue, a small tubal os, and decreased length and caliber of the tube.¹¹ Abnormal tubal anatomy caused by this exposure multiplies the risk of ectopic pregnancy by a factor of 5.⁹

In vitro fertilization. When blocked tubes are treated, the embryos can migrate retrograde into the oviduct, implant, and eventually rupture. The OR for ectopic pregnancy with assisted reproduction is 4.0.⁸

 

Cigarette smoking increases the likelihood of ectopic pregnancy 2.5 times,¹² probably by affecting ciliary action within the fallopian tubes.

Salpingitis isthmic nodosa is anatomic thickening of the proximal portion of the fallopian tubes with multiple lumen diverticula. It increases the risk of ectopic pregnancy 1.5 times, compared with age- and race-matched controls.¹³

Don’t depend solely on risk factors. Many ectopic pregnancies present without them.

Symptoms. Many ectopic pregnancies never produce symptoms; rather, they resolve spontaneously or are timely diagnosed and treated medically. Risk factors should therefore be examined in any woman in early pregnancy and investigated further if ectopic pregnancy is likely.

When symptoms do occur, they usually involve 1 or all of the classic triad: amenorrhea, irregular bleeding, and lower abdominal pain. In addition, syncope, shock, and pain radiating to the patient’s shoulder can result from hemoperitoneum.

TABLE 1

High, moderate, and low levels of risk factors for ectopic pregnancy

RISK FACTOR	ODDS RATIO*
High risk
Tubal surgery	21.0
Tubal ligation	9.3
Previous ectopic pregnancy	8.3
In utero exposure to diethylstilbestrol	5.6
Use of intrauterine device	4.2–45.0
Documented tubal pathology	3.8–21.0
Assisted reproduction	4.0
Emergency contraception	High
Moderate risk
Infertility	2.5–21.0
Previous genital infections	2.5–3.7
Multiple sexual partners	2.1
Salpingitis (isthmic)	1.5
Slight risk
Previous pelvic, abdominal surgery	0.9–3.8
Cigarette smoking	2.3–2.5
Vaginal douching	1.1–3.1
Early age at first intercourse (<18 years)	1.6
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).
* Single values = common odds ratio from homogeneous studies; point estimates = range of values from heterogeneous studies

STEP 2Document the pregnancy and measure ß-hCG

Once you identify the high-risk patient, or a woman comes in complaining of pain and spotting or bleeding, run a pregnancy test to confirm that she is pregnant and, if it is positive, obtain a quantitative ß-hCG.

ß-hCG levels are normally measured using enzyme-linked immunosorbent assays (ELISA), which detect ß-hCG in urine and serum at levels as low as 20 mIU/mL and 10 mIU/mL, respectively.¹⁴ ß-hCG is produced by trophoblastic cells in normal pregnancy, and approximately doubles every 2 days when titers are below 10,000 mIU/mL¹⁵—although in some normal pregnancies, ß-hCG may increase as slowly as 53% or as rapidly as 230% over 2 days.¹⁶ Eighty-five percent of abnormal pregnancies—whether intrauterine or ectopic—have impaired ß-hCG production with prolonged doubling time. Thus, in failing pregnancies, ß-hCG levels will plateau or fail to rise normally.

A single ß-hCG level fails to predict the risk of rupture, since ectopic pregnancies can rupture at ß-hCG levels as low as 10 mIU/mL or far exceeding 10,000 mIU/mL, or at any level in between.

STEP 3Obtain an ultrasound scan

Transvaginal ultrasound reliably detects normal intrauterine gestations when ß-hCG passes somewhere between 1,000 mIU/mL and 2,000 mIU/mL (First International Reference Preparation), depending on the expertise of the ultrasonographer and the particular equipment used.^8,17 This is known as the “discriminatory zone.” ß-hCG levels reach this zone as early as 1 week after missed menses.¹⁸

The discriminatory zone is not the lowest ß-hCG concentration at which an intrauterine pregnancy can be visualized via ultrasound. Rather, it is the value at which any intrauterine pregnancy will be apparent. At that value, the absence of an intrauterine pregnancy confirms—by negative conclusion—that the patient has a nonviable gestation.

When intrauterine pregnancy is visualized. The diagnosis is definitive and the woman’s symptoms can be explained as “threatened abortion.” No further investigation is necessary aside from routine prenatal care if the pregnancy continues.

When an extrauterine gestation is observed, such as a gestational sac with a detectable fetal heart rate, ectopic pregnancy can be diagnosed with 100% specificity but low sensitivity (15% to 20%). A complex adnexal mass without an intrauterine pregnancy improves sensitivity from 21% to 84% at the expense of lower specificity (93% to 99.5%).¹⁹

Even when an adnexal mass is visualized, cardiac activity is not usually present. If cardiac activity is apparent, proceed to surgery, since methotrexate usually will not resolve these gestations.

Be aware that some adnexal masses suspicious for ectopic pregnancy may turn out to be other entities, such as a corpus luteum, hydrosalpinx, ovarian neoplasm, or endometrioma. Unless a fetal heart rate is detected by ultrasound, the diagnosis is uncertain and curettage is needed to establish a definitive diagnosis.

No intrauterine pregnancy, no extrauterine mass. Despite the high resolution of transvaginal ultrasound, many patients with ectopic pregnancy have no apparent adnexal mass,²⁰ particularly when diagnosis is early. In these cases, proceed to curettage (step 4).

Don’t interpret ultrasound findings in a vacuum

This is especially unwise when ß-hCG levels are low—even when the ultrasound report points to intrauterine pregnancy. At ß-hCG levels below 1,500 mIU/mL, the sensitivity of ultrasound in diagnosing intrauterine pregnancy drops from 98% to 33% and predictive value is substantially lower. Interpret ultrasound and ß-hCG levels together for greater accuracy.

 

How size influences management

Ultrasound can detect ectopic pregnancies as small as 2 cm. In general, an ectopic sac size larger than 4 cm should be treated surgically.

STEP 4Perform uterine curettage

If ultrasound imaging is inconclusive and ß-hCG levels are plateauing or rising subnormally, perform uterine curettage. If ß-hCG levels decrease 15% or more 8 to 12 hours after the procedure, a complete abortion can be strongly suspected.²¹ If ß-hCG levels plateau or rise, the trophoblasts were not removed by curettage, and ectopic pregnancy is diagnosed.²¹

Keep in mind these important points:

• Without uterine curettage, roughly 40% of ectopic pregnancy diagnoses are incorrect.²²
• Because curettage will result in termination of pregnancy, it is vital that it be limited to cases involving abnormal ß-hCG levels, not normally rising values.

STEP 5Administer methotrexate

Medical management is indicated when the following circumstances are present:

• No viable intrauterine pregnancy is present.
• No rupture has occurred.
• Any adnexal mass is 4 cm in size or smaller.
• ß-hCG levels are below 10,000 mIU/mL.

If there is a positive fetal heart beat, surgery is preferred.

Firm diagnosis of ectopic pregnancy is essential prior to methotrexate administration. If the drug is given to a woman carrying a viable pregnancy, it may result in loss of the pregnancy or methotrexate embryopathy.²³

How methotrexate resolves ectopic pregnancy

Systemic methotrexate has been used successfully in the treatment of ectopic pregnancy for more than 20 years. A folic acid antagonist, it inhibits de novo synthesis of purines and pyrimidines. It thus interferes with DNA synthesis and cell multiplication.²⁴ Actively proliferating trophoblasts are particularly vulnerable.²⁵

When methotrexate is administered to normally pregnant women, it blunts the normal ß-hCG increment over the next 7 days. Circulating progesterone and 17-a-hydroxyprogesterone concentrations also decline, and abortion occurs.²⁶

Methotrexate directly impairs trophoblastic production of hCG; the decrement in corpus luteum progesterone is a secondary event.

Side effects include abdominal distress, chills and fever, dizziness, immunosuppression, leukopenia, malaise, nausea, ulcerative stomatitis, photosensitivity, and undue fatigue.

Breastfeeding is an absolute contraindication to methotrexate, while relative contraindications include abnormal liver function tests, blood dyscrasias, excessive alcohol consumption, HIV/AIDS, psoriasis, ongoing radiotherapy, rheumatoid arthritis, and significant pulmonary disease.

Multiple-dose methotrexate is superior

In a recent meta-analysis comparing single-and multiple-dose regimens of methotrexate, Barnhart and colleagues⁴ found the latter to be more effective, although the single dose was more commonly given. The single-dose regimen was associated with a significantly greater chance of failure in both crude (OR 1.71; 1.04, 2.82) and adjusted (OR 4.74; 1.77, 12.62) analyses. Consequently, we advocate the multiple-dose regimen detailed in TABLE 2: 1 mg of methotrexate per kilogram of body weight on day 1, alternating with 0.1 mg of leucovorin per kilogram on succeeding days. Continue this regimen until ß-hCG levels decline in 2 consecutive daily titers, or 4 doses of methotrexate are given, whichever comes first.

If treatment is unsuccessful after 4 doses, additional methotrexate is unlikely to be effective and will involve significant additional cost and morbidity. If ß-hCG levels plateau or continue to rise, surgery is indicated.

Although multiple-dose therapy is more effective than a single dose, the optimal number of doses probably falls somewhere between 1 and 4.⁴ A 2-dose protocol currently under investigation may provide an optimal compromise.

Artificially high efficacy rates with “single-dose” therapy. Unusually high success rates in initial studies of the single-dose regimen may have been due to the inclusion of spontaneously aborting intrauterine pregnancies.²⁷ Although 6 subsequent studies—1 cohort and 5 case-control studies involving 304 patients—found an overall success rate (no surgical intervention) of 87.2%, 11.5% of participants required more than 1 dose.⁹

Still the standard. Despite its lower efficacy rates, single-dose methotrexate remains the “standard” in the United States, as recommended in an American College of Obstetricians and Gynecologists practice bulletin.²⁸ The usual intramuscular dose is 50 mg per square meter of body surface area, with ß-hCG titers on days 4 and 7 and an additional dose if ß-hCG levels fail to resolve.

What the evidence shows

Multidose methotrexate. Twelve studies measured the success of multiple-dose systemic methotrexate (TABLE 3); they included 1 randomized, controlled trial, 1 cohort study, and 10 case series. Between 1982 and 1997, 325 cases were treated with multiple-dose methotrexate. Of these, 93.8% had successful resolution with no subsequent therapy, and 78.9% of the 161 women tested had patent oviducts. In addition, of 95 women hoping to conceive, 57.9% had a subsequent intrauterine pregnancy and 7.4% developed a recurrent ectopic pregnancy. These rates compared favorably with numerous laparoscopic surgery series published during the same years.⁹

 

Methotrexate versus laparoscopic salpingostomy. In 1 randomized clinical comparison,²⁹ 100 women with laparoscopy-confirmed ectopic pregnancy were randomized to methotrexate or laparoscopic salpingostomy. Of the 51 patients treated medically, 7 (14%) required surgical intervention for active bleeding and/or tubal rupture. An additional course of methotrexate was required in 2 patients (4%) for persistent trophoblasts.

Of the 49 patients in the salpingostomy group, 4 women (8%) failed therapy and required salpingectomies, and 10 patients (20%) were treated with methotrexate for persistent trophoblasts.

Tubal patency was present in 23 of 42 women (55%) in the methotrexate group, compared with 23 of 39 (59%) in the salpingostomy group.

Overall, this randomized study²⁹ and previous meta-analysis demonstrate that systemic multiple-dose methotrexate is comparable in efficacy to laparoscopic salpingostomy.

TABLE 2

Multiple-dose methotrexate protocol

Discontinue treatment when there is a decline in 2 consecutive ß-hCG titers or after 4 doses, whichever comes first.

DAY	INTERVENTION	DOSE (MG)
1	Baseline studies ß-hCG titer, CBC, and platelets Methotrexate	1.0
2	Leucovori	0.1
3	Methotrexate	1.0
4	Leucovorin ß-hCG titer	0.1
5	Methotrexate ß-hCG titer	1.0
6	Leucovorin ß-hCG titer	0.1
7	Methotrexate ß-hCG titer	1.0
8	Leucovorin ß-hCG titer CBC and platelets Renal and liver function tests	0.1
Weekly	ß-hCG titer until negative

TABLE 3

Treatment outcomes for ectopic pregnancy

METHOD	NUMBER OF STUDIES	NUMBER OF PATIENTS	NUMBER WITH SUCCESSFUL RESOLUTION	TUBAL PATENCY RATE	SUBSEQUENT FERTILITY RATE
					INTRAUTERINE PREGNANCY	ECTOPIC PREGNANCY
Conservative laparoscopic surgery	32	1,626	1,516 (93%)	170/223 (76%)	366/647 (57%)	87/647 (13%)
Variable-dose methotrexate	12	338	314 (93%)	136/182 (75%)	55/95 (58%)	7/95 (7%)
Single-dose methotrexate	7	393	340 (87%)	61/75 (81%)	39/64 (61%)	5/64 (8%)
Direct-injection methotrexate	21	660	502 (76%)	130/162 (80%)	87/152 (57%)	9/152 (6%)
Expectant management	14	628	425 (68%)	60/79 (76%)	12/14 (86%)	1/14 (7%)
Reprinted with permission from Elsevier (The Lancet, 1998, vol 351, 1115–1120).

Fine points of treatment

During methotrexate therapy, examine the patient only once to avoid triggering a rupture, and counsel her to avoid intercourse for the same reason.³⁰ Do not perform repeat vaginal ultrasound examination. Also inform her that transient pain (“separation pain”) from tubal abortion frequently occurs 3 to 7 days after the start of therapy, lasts 4 to 12 hours, and then resolves.³¹ This is perhaps the most difficult aspect of methotrexate therapy, as it is not always easy to differentiate the pain of tubal abortion from the pain of rupture.

If ß-hCG titers continue to rise rapidly between methotrexate doses, rupture is more likely and surgery should proceed.³²

Overall, isthmic ectopic pregnancies (12.3% of ectopics) appear to be at a particularly high risk for rupture and comprise nearly half of methotrexate failures.³² Unfortunately, there is no way to identify isthmic pregnancies without surgery.

Avoid NSAIDs and GI-“unfriendly” foods. Counsel the patient to avoid nonsteroidal anti-inflammatory drugs (NSAIDs) because they may impair natural hemostasis. Gasforming foods such as leeks, corn, and cabbage can cause distension, which may be mistaken for rupture.

Also instruct the patient to avoid folic acid, which impairs the efficacy of methotrexate.

Ultrasound surveillance is unnecessary. If an adnexal mass was identified at initial imaging, there is no need to view it again, since these masses tend to enlarge and form hematomas and can cause undue anxiety in both physician and patient. In properly selected patients, multidose methotrexate with monitoring of ß-hCG levels should suffice.

When surgery is indicated

Surgical intervention is necessary when pain is severe, persists beyond 12 hours, and is associated with orthostatic hypotension, falling hematocrit, or persistently elevated ß-hCG levels after methotrexate therapy.

Laparoscopy is the preferred approach. Advantages include less blood loss and analgesia,³³ shortened postoperative recovery, and lower costs.

Technique. For unruptured ampullary ectopic pregnancy, salpingostomy is preferred. Make a linear incision over the bulging antimesenteric border of the fallopian tube using electrocautery, scissors, or laser. Remove the products of conception using forceps or suction, and leave the incision to heal by secondary intention.

For isthmic pregnancies, use segmental excision followed by delayed microsurgical anastomosis.³⁴ The isthmic tubal lumen is narrower and the muscularis thicker than in the ampulla. Thus, the isthmus is predisposed to greater damage after salpingostomy and greater rates of proximal obstruction.³³

An increased risk of persistent ectopic pregnancy has been a criticism of salpingostomy. However, when 1 dose of systemic methotrexate is combined with salpingostomy, the risk of persistent pregnancy is virtually eliminated.³⁵

When rupture occurs, salpingectomy is the first choice for treatment, as it arrests hemorrhage and shortens the procedure. Either laparoscopy or laparotomy is appropriate.

The authors report no financial relationships relevant to this article.

CASE 1

A 42-year-old woman of Ashkenazi Jewish ancestry, whose mother had a diagnosis of ovarian cancer a year earlier, is worried that she may also be at risk.

Her family has experienced no other cases of breast, colorectal, or ovarian cancer. She has 2 children and has used oral contraceptives for 9 years.

Primary prevention is an alternative to diagnosis of ovarian cancer at an early stage—a goal that is all too often unattainable.

If we identify women with increased risk based on their family history, environmental exposure, hormone use, or reproductive experiences, we can counsel them about lifestyle changes, chemoprophylaxis, or preventive surgery, depending on their reproductive desires and overall ovarian cancer risk.

Using 2 hypothetical cases, this article describes a quick history to screen for potential high risk, and summarizes what we can advise concerned patients, based on findings to date.

Topics include:

• genetic risk assessment
  
• when to refer to a genetic counselor
  
• risk factors
  
• oral contraceptives for primary prevention
  
• prophylactic surgery
  

Despite advances in medical and surgical therapies, ovarian cancer remains the deadliest gynecologic malignancy and the fifth leading cause of cancer deaths among women in the United States. This year the disease will strike an estimated 25,400 women in this country and kill more than 14,000.¹

Early detection versus primary prevention

One of the biggest problems is well known: In its early stages, ovarian cancer often is asymptomatic. Even when advanced, symptoms tend to be vague and are often dismissed by patients and their doctors. By the time the diagnosis is suspected, most women have disease beyond the ovary.

Because curability greatly depends on stage at presentation, much research is directed toward early detection. Advances in high-resolution imaging and novel blood or urine tumor markers may one day offer effective screening, but at present these methods are unproven.

3 genetic syndromes

Three genetic syndromes account for the vast majority of familial ovarian cancer and approximately 10% of all ovarian cancers. They are:

• breast-ovarian cancer syndrome,
  
• site-specific ovarian cancer syndrome, and
  
• hereditary nonpolyposis colorectal cancer syndrome (HNPCC) (Lynch II).²
  

The first 2 are caused by inherited mutations in the BRCA1 and BRCA2 genes. In fact, though often described as separate entities, these syndromes are likely phenotypic variants of the same genetic mutations. BRCA1 and BRCA2 function as classic tumor-suppressor genes and are inherited in an autosomal dominant fashion.

HNPCC is caused by mutations in a series of genes responsible for repairing errors in DNA replication. Inactivation of these so-called mismatch repair genes results in a high incidence of right-sided colon cancer, endometrial cancer, and ovarian cancer.³

Gene mutations in different populations

The lifetime risk of developing ovarian cancer in the United States is about 1.4%. However, among women with BRCA1 or BRCA2 mutations, the risk rises to 20% to 60%.⁴

These genes also impart a significant lifetime risk of breast cancer in women and, in the case of BRCA2, in men as well.

Less than 0.15% of the general population carries BRCA1 or BRCA2 mutations. However, the carrier rate is dependent on ethnic background.⁵ Founder mutations have been identified among multiple unrelated families in Iceland, the Netherlands, and Sweden, and among Jews of Central or Eastern European descent (Ashkenazi).

The best described founder mutations are the 185delAG and 5382insC mutations in BRCA1 and the 6174delT mutation in BRCA2, occurring in Ashkenazi Jews at a carrier rate of 2%.⁶

Ovarian cancer in Ashkenazi women more likely genetic than sporadic

Although an Ashkenazi woman is no more likely to develop ovarian cancer than a noncarrier, if she does develop the disease, it is far more likely to be genetic rather than sporadic.

Consequently, if a woman of Ashkenazi Jewish descent develops ovarian cancer, there is a 40% chance she carries a mutation in one of these 2 genes.⁷ Her first-degree relatives (mother, sisters, daughters) have a 20% risk of being gene carriers (50% in autosomal dominant transmission).

Therefore, an Ashkenazi Jewish woman needs only 1 first-degree relative with ovarian cancer to be considered for further genetic counseling.

 

The 3-2-1 rule for genetic counseling

Genetic counseling is indicated when the patient meets the “3-2-1” rule, known as the modified Amsterdam criteria:

• 3 affected individuals with either colorectal or ovarian cancer, in
  
• 2 successive generations, with at least
  
• 1 who developed cancer under the age of 50 years.
  

An Ashkenazi woman with only 1 first-degree relative with ovarian cancer is eligible for genetic counseling.

When HNPCC is present

Women with documented HNPCC have a 70% lifetime risk of developing endometrial cancer and a lifetime risk of ovarian cancer of 11% or more. Testing for mutation in mismatch repair genes can be performed on peripheral leukocytes. Alternatively, the primary tumor from affected individuals can be assessed for the presence of microsatellite instability, a consequence of defective mismatch repair.³

How to obtain a family history

Women with a significant family history of breast, colorectal, endometrial, or ovarian cancer may face an elevated risk of ovarian cancer. Basic questions about the occurrence of these malignancies in first-degree (mother, sister, daughter) or second-degree (aunt, grandmother) relatives must be a component of every gynecologic history.

1-question assessment

The following question is a simple but effective way to assess familial cancer risk: “Has anyone in your family had breast cancer under the age of 35, or colorectal, uterine, or ovarian cancer?” If so, ask further questions about the relative’s age at onset and how the affected person is related to the patient.

Although an informal pedigree can be constructed in a few moments, a formal pedigree is a more daunting task. More than half of family histories of ovarian cancer are inaccurate—and the error rate increases if the affected family member is a distant relative.⁸

Moreover, not all ovarian cancers are associated with genetic syndromes. Serous epithelial ovarian, fallopian tube, and peritoneal cancers dominate in BRCA-associated syndromes, whereas mucinous epithelial cancers are very rare. Germ-cell and stromal neoplasms as well as epithelial tumors of low malignant potential are excluded when assessing genetic risk.

Additional documentation is often necessary to verify the history. Such medical detective work may be beyond what a generalist is willing to perform. In that case, referral to a genetic counselor is the next step.

When to refer

Recommendation. If she is of Ashkenazi Jewish descent, like the patient described at the beginning of this article, only 1 first-degree relative indicates referral for genetic counseling.

If the patient claims at least 2 first-degree or second-degree relatives with breast cancer under the age of 35, or with ovarian cancer, suspect a genetic syndrome.

Family histories of particular concern include breast and ovarian cancer in a single individual, or any case of male breast cancer. Refer these patients to genetic counseling for formal pedigree analysis and possible testing for genetic predisposition. In the case of HNPCC, refer the patient if the modified Amsterdam criteria are met.

What does a genetic counselor do?

One of the most important roles of a genetic counselor is constructing as accurate a pedigree as possible. In some cases, this requires much effort on the part of both patient and counselor. It is not uncommon for the counselor to request outside pathology reports, operative notes, and death certificates to verify cancer cases. Pathology slides also may be necessary, while tissue blocks occasionally are requested to assist in genetic testing.

When testing is warranted

Genetic predisposition testing should be undertaken only after formal consultation with a genetic counselor who has expertise in cancer predisposition syndromes. As a policy statement from the American Society of Clinical Oncology concludes, genetic testing for cancer susceptibility should be performed only when:

Obesity is associated with increased risk of ovarian cancer mortality.

• pre- and post-test counseling is available,
  
• the test can be interpreted, and
  
• the results will help in medical and surgical management.
  

Ideally, testing should take place in the setting of a multidisciplinary team with expertise in the interpretation of verified family cancer pedigrees and in the medical, emotional, financial, and legal ramifications of genetic testing.⁹

CASE 2

A 65-year-old obese nullipara is currently on year 10 of combined hormone replacement therapy (HRT). Although she has no family history of breast or ovarian cancer, she asks about her risk of developing ovarian cancer.

Other risk factors

Findings on the following potential risk factors for ovarian cancer may help address this patient’s concerns:

Dietary factors

In a recent study, obesity was associated with an increased risk of ovarian cancer mortality.¹⁰ Women who eat a diet high in saturated fat and low in vegetable fiber also may face an increased risk.

 

In 1989, the observation that Swedish women had both a high risk of ovarian cancer and the highest per capita dairy consumption in the world led some investigators to postulate a relationship between lactose consumption and ovarian cancer. The reason: When compared with matched controls, women with ovarian cancer were more likely to have high levels of galactose, a component sugar of the disaccharide lactose and a known oocyte toxin.¹¹

This observation, however, has been inconsistent. Therefore, no specific dietary strategy can be recommended to reduce the risk of ovarian cancer.

Recommendation. I would advise this patient to maintain a normal body mass index. Data are insufficient to support more specific diet recommendations.

Talc exposure

When talc is placed on the perineum, it may enter the vagina and ascend to the upper genital tract. Because talc is structurally similar to asbestos, it may theoretically increase the ovarian cancer risk. The observation that women who undergo tubal sterilization procedures or hysterectomy have a lower risk of ovarian cancer supports the ascending-carcinogen hypothesis.

Multiple case-control studies have shown a small but consistent increased risk with perineal application of talc (odds ratio 1.3, 95% confidence interval [CI], 1.1–1.6).¹² The risk appears to be time- and dose-dependent, with greater risk associated with more frequent application of perineal talc over a long duration.

Recommendation. The practice of applying genuine talc to the perineum should be discouraged. Cornstarch-based dusting powders are widely available.

Infertility drugs

One of the most difficult issues to study is the relationship between infertility drugs and ovarian cancer, although we know that unexplained infertility is an independent risk factor for ovarian cancer.

A retrospective study that claimed an association between prolonged clomiphene exposure and ovarian cancer¹³ was not restricted to invasive epithelial ovarian cancers, but included granulosa cell tumors. These estrogen-secreting neoplasms of stromal origin may directly contribute to infertility by disrupting normal follicular maturation and the menstrual cycle.

A number of studies—including a large collaborative analysis of 12 case-control studies—have reported an association between fertility drugs and invasive epithelial ovarian cancer.¹⁴ In addition, many of the theoretical models of epithelial ovarian cancer pathogenesis implicate both incessant ovulation and high gonadotropin levels as important steps in malignant transformation of ovarian epithelium.

Oral contraceptives (OCs), which reduce ovulatory events, and moderate gonadotropin levels are associated with a consistent and significant protective effect.

Recommendation. It seems prudent, in the absence of convincing data, to use fertility medication only when absolutely indicated, at the lowest effective dose, and for the shortest duration possible without compromising treatment success.

However, prior exposure to these agents should not be considered an indication for increased surveillance or prophylactic surgery.

Estrogen replacement therapy

Women on estrogen replacement therapy appear to have an increased risk of ovarian cancer. When compared to nonusers, “ever-users” had a relative risk of ovarian cancer of 2.² (95% CI, 1.53–3.17), and the risk increased with the duration of use.¹⁵ Long-term users, defined as women who used estrogen replacement therapy for at least 20 years, had a relative risk of 3.2 (95% CI, 1.7–5.7).¹⁶

Although some studies suggest a protective effect of combination hormone replacement regimens that include both estrogen and progesterone, this observation has not been confirmed. Thus, long-term estrogen users should consider an increased risk of developing ovarian cancer when deciding whether to initiate or continue estrogen replacement therapy.

Recommendation. I would advise the Case 2 patient to stop HRT.

Primary prevention

Oral contraceptives

OCs reduce the risk of ovarian cancer significantly. A number of studies have demonstrated a 10% risk reduction per year for up to 5 to 7 years of use.¹⁷ This effect appears to persist for at least 10 years after OCs are discontinued. It also has been observed in patients known to be BRCA1 and BRCA2 carriers and is the basis for recommending OCs as a chemoprophylactic method in known carriers who wish to retain fertility.¹⁸

Long-term estrogen users should consider increased risk of ovarian cancer. Ovarian cancer: High-risk patients.

However, use of OCs by BRCA carriers is not without some controversy. An Israeli population-based study¹⁹ of ovarian cancer and OC use demonstrated a protective effect of pregnancy but not use of OCs. It is unclear why the Israeli data is inconsistent with prior published reports.

Recommendation. In Case 1, the patient has taken OCs for 9 years. Her risk has been substantially reduced, and I would advise her to continue OC pills when not attempting to get pregnant, until she is near menopause.

Prophylactic surgery

In high-risk women, preventive surgery may substantially reduce but not completely eliminate the risk of ovarian cancer. For example, bilateral salpingo-oophorectomy in BRCA carriers reduces the risk of ovarian cancer by more than 90% and the risk of breast cancer by more than 50%.^20,21

 

The operation should be reserved for women with known mutations in BRCA1 or BRCA2 or who have a family history consistent with one of the genetic syndromes associated with ovarian cancer.

The addition of hysterectomy does not appear to increase the efficacy of the operation and should be performed only for concurrent gynecologic indications or if the patient has HNPCC.

Patients should be informed that prophylactic surgery does not protect against subsequent papillary serous carcinoma of the peritoneum. They also should be warned that about 7% of operations detect occult ovarian or tubal carcinoma, which may not be identified until final pathology reports are issued.²² Pathologists should be instructed to submit the entire specimen for sectioning to reduce risk of missing microscopic occult malignancy.

In addition, the patient should be prepared for surgical menopause.

Effective primary prevention strategies such as chemoprophylaxis and prophylactic surgery, when appropriately applied, may spare many women the devastating consequences of this dreaded disease.

Internet resources

• Ovarian Cancer (PDQ): Prevention of ovarian cancer http://www.cancer.gov/cancerinfo/pdq/prevention/ovarian/healthprofessional
  
• Myriad Genetics: BRCA1/2 mutation prevalence tables http:///www.myriadtests.com/provider/mutprevo.htm
  
• National Society of Genetic Counselors: Locate a genetic counselor specializing in cancer risk assessment http://www.nsgc.org/resourcelink.asp
  

The author reports no financial relationships relevant to this article.

CASE 1

A 42-year-old woman of Ashkenazi Jewish ancestry, whose mother had a diagnosis of ovarian cancer a year earlier, is worried that she may also be at risk.

Her family has experienced no other cases of breast, colorectal, or ovarian cancer. She has 2 children and has used oral contraceptives for 9 years.

Primary prevention is an alternative to diagnosis of ovarian cancer at an early stage—a goal that is all too often unattainable.

If we identify women with increased risk based on their family history, environmental exposure, hormone use, or reproductive experiences, we can counsel them about lifestyle changes, chemoprophylaxis, or preventive surgery, depending on their reproductive desires and overall ovarian cancer risk.

Using 2 hypothetical cases, this article describes a quick history to screen for potential high risk, and summarizes what we can advise concerned patients, based on findings to date.

Topics include:

• genetic risk assessment
  
• when to refer to a genetic counselor
  
• risk factors
  
• oral contraceptives for primary prevention
  
• prophylactic surgery
  

Despite advances in medical and surgical therapies, ovarian cancer remains the deadliest gynecologic malignancy and the fifth leading cause of cancer deaths among women in the United States. This year the disease will strike an estimated 25,400 women in this country and kill more than 14,000.¹

Early detection versus primary prevention

One of the biggest problems is well known: In its early stages, ovarian cancer often is asymptomatic. Even when advanced, symptoms tend to be vague and are often dismissed by patients and their doctors. By the time the diagnosis is suspected, most women have disease beyond the ovary.

Because curability greatly depends on stage at presentation, much research is directed toward early detection. Advances in high-resolution imaging and novel blood or urine tumor markers may one day offer effective screening, but at present these methods are unproven.

3 genetic syndromes

Three genetic syndromes account for the vast majority of familial ovarian cancer and approximately 10% of all ovarian cancers. They are:

• breast-ovarian cancer syndrome,
  
• site-specific ovarian cancer syndrome, and
  
• hereditary nonpolyposis colorectal cancer syndrome (HNPCC) (Lynch II).²
  

The first 2 are caused by inherited mutations in the BRCA1 and BRCA2 genes. In fact, though often described as separate entities, these syndromes are likely phenotypic variants of the same genetic mutations. BRCA1 and BRCA2 function as classic tumor-suppressor genes and are inherited in an autosomal dominant fashion.

HNPCC is caused by mutations in a series of genes responsible for repairing errors in DNA replication. Inactivation of these so-called mismatch repair genes results in a high incidence of right-sided colon cancer, endometrial cancer, and ovarian cancer.³

Gene mutations in different populations

The lifetime risk of developing ovarian cancer in the United States is about 1.4%. However, among women with BRCA1 or BRCA2 mutations, the risk rises to 20% to 60%.⁴

These genes also impart a significant lifetime risk of breast cancer in women and, in the case of BRCA2, in men as well.

Less than 0.15% of the general population carries BRCA1 or BRCA2 mutations. However, the carrier rate is dependent on ethnic background.⁵ Founder mutations have been identified among multiple unrelated families in Iceland, the Netherlands, and Sweden, and among Jews of Central or Eastern European descent (Ashkenazi).

The best described founder mutations are the 185delAG and 5382insC mutations in BRCA1 and the 6174delT mutation in BRCA2, occurring in Ashkenazi Jews at a carrier rate of 2%.⁶

Ovarian cancer in Ashkenazi women more likely genetic than sporadic

Although an Ashkenazi woman is no more likely to develop ovarian cancer than a noncarrier, if she does develop the disease, it is far more likely to be genetic rather than sporadic.

Consequently, if a woman of Ashkenazi Jewish descent develops ovarian cancer, there is a 40% chance she carries a mutation in one of these 2 genes.⁷ Her first-degree relatives (mother, sisters, daughters) have a 20% risk of being gene carriers (50% in autosomal dominant transmission).

Therefore, an Ashkenazi Jewish woman needs only 1 first-degree relative with ovarian cancer to be considered for further genetic counseling.

 

The 3-2-1 rule for genetic counseling

Genetic counseling is indicated when the patient meets the “3-2-1” rule, known as the modified Amsterdam criteria:

• 3 affected individuals with either colorectal or ovarian cancer, in
  
• 2 successive generations, with at least
  
• 1 who developed cancer under the age of 50 years.
  

An Ashkenazi woman with only 1 first-degree relative with ovarian cancer is eligible for genetic counseling.

When HNPCC is present

Women with documented HNPCC have a 70% lifetime risk of developing endometrial cancer and a lifetime risk of ovarian cancer of 11% or more. Testing for mutation in mismatch repair genes can be performed on peripheral leukocytes. Alternatively, the primary tumor from affected individuals can be assessed for the presence of microsatellite instability, a consequence of defective mismatch repair.³

How to obtain a family history

Women with a significant family history of breast, colorectal, endometrial, or ovarian cancer may face an elevated risk of ovarian cancer. Basic questions about the occurrence of these malignancies in first-degree (mother, sister, daughter) or second-degree (aunt, grandmother) relatives must be a component of every gynecologic history.

1-question assessment

The following question is a simple but effective way to assess familial cancer risk: “Has anyone in your family had breast cancer under the age of 35, or colorectal, uterine, or ovarian cancer?” If so, ask further questions about the relative’s age at onset and how the affected person is related to the patient.

Although an informal pedigree can be constructed in a few moments, a formal pedigree is a more daunting task. More than half of family histories of ovarian cancer are inaccurate—and the error rate increases if the affected family member is a distant relative.⁸

Moreover, not all ovarian cancers are associated with genetic syndromes. Serous epithelial ovarian, fallopian tube, and peritoneal cancers dominate in BRCA-associated syndromes, whereas mucinous epithelial cancers are very rare. Germ-cell and stromal neoplasms as well as epithelial tumors of low malignant potential are excluded when assessing genetic risk.

Additional documentation is often necessary to verify the history. Such medical detective work may be beyond what a generalist is willing to perform. In that case, referral to a genetic counselor is the next step.

When to refer

Recommendation. If she is of Ashkenazi Jewish descent, like the patient described at the beginning of this article, only 1 first-degree relative indicates referral for genetic counseling.

If the patient claims at least 2 first-degree or second-degree relatives with breast cancer under the age of 35, or with ovarian cancer, suspect a genetic syndrome.

Family histories of particular concern include breast and ovarian cancer in a single individual, or any case of male breast cancer. Refer these patients to genetic counseling for formal pedigree analysis and possible testing for genetic predisposition. In the case of HNPCC, refer the patient if the modified Amsterdam criteria are met.

What does a genetic counselor do?

One of the most important roles of a genetic counselor is constructing as accurate a pedigree as possible. In some cases, this requires much effort on the part of both patient and counselor. It is not uncommon for the counselor to request outside pathology reports, operative notes, and death certificates to verify cancer cases. Pathology slides also may be necessary, while tissue blocks occasionally are requested to assist in genetic testing.

When testing is warranted

Genetic predisposition testing should be undertaken only after formal consultation with a genetic counselor who has expertise in cancer predisposition syndromes. As a policy statement from the American Society of Clinical Oncology concludes, genetic testing for cancer susceptibility should be performed only when:

Obesity is associated with increased risk of ovarian cancer mortality.

• pre- and post-test counseling is available,
  
• the test can be interpreted, and
  
• the results will help in medical and surgical management.
  

Ideally, testing should take place in the setting of a multidisciplinary team with expertise in the interpretation of verified family cancer pedigrees and in the medical, emotional, financial, and legal ramifications of genetic testing.⁹

CASE 2

A 65-year-old obese nullipara is currently on year 10 of combined hormone replacement therapy (HRT). Although she has no family history of breast or ovarian cancer, she asks about her risk of developing ovarian cancer.

Other risk factors

Findings on the following potential risk factors for ovarian cancer may help address this patient’s concerns:

Dietary factors

In a recent study, obesity was associated with an increased risk of ovarian cancer mortality.¹⁰ Women who eat a diet high in saturated fat and low in vegetable fiber also may face an increased risk.

 

In 1989, the observation that Swedish women had both a high risk of ovarian cancer and the highest per capita dairy consumption in the world led some investigators to postulate a relationship between lactose consumption and ovarian cancer. The reason: When compared with matched controls, women with ovarian cancer were more likely to have high levels of galactose, a component sugar of the disaccharide lactose and a known oocyte toxin.¹¹

This observation, however, has been inconsistent. Therefore, no specific dietary strategy can be recommended to reduce the risk of ovarian cancer.

Recommendation. I would advise this patient to maintain a normal body mass index. Data are insufficient to support more specific diet recommendations.

Talc exposure

When talc is placed on the perineum, it may enter the vagina and ascend to the upper genital tract. Because talc is structurally similar to asbestos, it may theoretically increase the ovarian cancer risk. The observation that women who undergo tubal sterilization procedures or hysterectomy have a lower risk of ovarian cancer supports the ascending-carcinogen hypothesis.

Multiple case-control studies have shown a small but consistent increased risk with perineal application of talc (odds ratio 1.3, 95% confidence interval [CI], 1.1–1.6).¹² The risk appears to be time- and dose-dependent, with greater risk associated with more frequent application of perineal talc over a long duration.

Recommendation. The practice of applying genuine talc to the perineum should be discouraged. Cornstarch-based dusting powders are widely available.

Infertility drugs

One of the most difficult issues to study is the relationship between infertility drugs and ovarian cancer, although we know that unexplained infertility is an independent risk factor for ovarian cancer.

A retrospective study that claimed an association between prolonged clomiphene exposure and ovarian cancer¹³ was not restricted to invasive epithelial ovarian cancers, but included granulosa cell tumors. These estrogen-secreting neoplasms of stromal origin may directly contribute to infertility by disrupting normal follicular maturation and the menstrual cycle.

A number of studies—including a large collaborative analysis of 12 case-control studies—have reported an association between fertility drugs and invasive epithelial ovarian cancer.¹⁴ In addition, many of the theoretical models of epithelial ovarian cancer pathogenesis implicate both incessant ovulation and high gonadotropin levels as important steps in malignant transformation of ovarian epithelium.

Oral contraceptives (OCs), which reduce ovulatory events, and moderate gonadotropin levels are associated with a consistent and significant protective effect.

Recommendation. It seems prudent, in the absence of convincing data, to use fertility medication only when absolutely indicated, at the lowest effective dose, and for the shortest duration possible without compromising treatment success.

However, prior exposure to these agents should not be considered an indication for increased surveillance or prophylactic surgery.

Estrogen replacement therapy

Women on estrogen replacement therapy appear to have an increased risk of ovarian cancer. When compared to nonusers, “ever-users” had a relative risk of ovarian cancer of 2.² (95% CI, 1.53–3.17), and the risk increased with the duration of use.¹⁵ Long-term users, defined as women who used estrogen replacement therapy for at least 20 years, had a relative risk of 3.2 (95% CI, 1.7–5.7).¹⁶

Although some studies suggest a protective effect of combination hormone replacement regimens that include both estrogen and progesterone, this observation has not been confirmed. Thus, long-term estrogen users should consider an increased risk of developing ovarian cancer when deciding whether to initiate or continue estrogen replacement therapy.

Recommendation. I would advise the Case 2 patient to stop HRT.

Primary prevention

Oral contraceptives

OCs reduce the risk of ovarian cancer significantly. A number of studies have demonstrated a 10% risk reduction per year for up to 5 to 7 years of use.¹⁷ This effect appears to persist for at least 10 years after OCs are discontinued. It also has been observed in patients known to be BRCA1 and BRCA2 carriers and is the basis for recommending OCs as a chemoprophylactic method in known carriers who wish to retain fertility.¹⁸

Long-term estrogen users should consider increased risk of ovarian cancer. Ovarian cancer: High-risk patients.

However, use of OCs by BRCA carriers is not without some controversy. An Israeli population-based study¹⁹ of ovarian cancer and OC use demonstrated a protective effect of pregnancy but not use of OCs. It is unclear why the Israeli data is inconsistent with prior published reports.

Recommendation. In Case 1, the patient has taken OCs for 9 years. Her risk has been substantially reduced, and I would advise her to continue OC pills when not attempting to get pregnant, until she is near menopause.

Prophylactic surgery

In high-risk women, preventive surgery may substantially reduce but not completely eliminate the risk of ovarian cancer. For example, bilateral salpingo-oophorectomy in BRCA carriers reduces the risk of ovarian cancer by more than 90% and the risk of breast cancer by more than 50%.^20,21

 

The operation should be reserved for women with known mutations in BRCA1 or BRCA2 or who have a family history consistent with one of the genetic syndromes associated with ovarian cancer.

The addition of hysterectomy does not appear to increase the efficacy of the operation and should be performed only for concurrent gynecologic indications or if the patient has HNPCC.

Patients should be informed that prophylactic surgery does not protect against subsequent papillary serous carcinoma of the peritoneum. They also should be warned that about 7% of operations detect occult ovarian or tubal carcinoma, which may not be identified until final pathology reports are issued.²² Pathologists should be instructed to submit the entire specimen for sectioning to reduce risk of missing microscopic occult malignancy.

In addition, the patient should be prepared for surgical menopause.

Effective primary prevention strategies such as chemoprophylaxis and prophylactic surgery, when appropriately applied, may spare many women the devastating consequences of this dreaded disease.

Internet resources

• Ovarian Cancer (PDQ): Prevention of ovarian cancer http://www.cancer.gov/cancerinfo/pdq/prevention/ovarian/healthprofessional
  
• Myriad Genetics: BRCA1/2 mutation prevalence tables http:///www.myriadtests.com/provider/mutprevo.htm
  
• National Society of Genetic Counselors: Locate a genetic counselor specializing in cancer risk assessment http://www.nsgc.org/resourcelink.asp
  

The author reports no financial relationships relevant to this article.

CASE 1

A 42-year-old woman of Ashkenazi Jewish ancestry, whose mother had a diagnosis of ovarian cancer a year earlier, is worried that she may also be at risk.

Her family has experienced no other cases of breast, colorectal, or ovarian cancer. She has 2 children and has used oral contraceptives for 9 years.

Primary prevention is an alternative to diagnosis of ovarian cancer at an early stage—a goal that is all too often unattainable.

If we identify women with increased risk based on their family history, environmental exposure, hormone use, or reproductive experiences, we can counsel them about lifestyle changes, chemoprophylaxis, or preventive surgery, depending on their reproductive desires and overall ovarian cancer risk.

Using 2 hypothetical cases, this article describes a quick history to screen for potential high risk, and summarizes what we can advise concerned patients, based on findings to date.

Topics include:

• genetic risk assessment
  
• when to refer to a genetic counselor
  
• risk factors
  
• oral contraceptives for primary prevention
  
• prophylactic surgery
  

Despite advances in medical and surgical therapies, ovarian cancer remains the deadliest gynecologic malignancy and the fifth leading cause of cancer deaths among women in the United States. This year the disease will strike an estimated 25,400 women in this country and kill more than 14,000.¹

Early detection versus primary prevention

One of the biggest problems is well known: In its early stages, ovarian cancer often is asymptomatic. Even when advanced, symptoms tend to be vague and are often dismissed by patients and their doctors. By the time the diagnosis is suspected, most women have disease beyond the ovary.

Because curability greatly depends on stage at presentation, much research is directed toward early detection. Advances in high-resolution imaging and novel blood or urine tumor markers may one day offer effective screening, but at present these methods are unproven.

3 genetic syndromes

Three genetic syndromes account for the vast majority of familial ovarian cancer and approximately 10% of all ovarian cancers. They are:

• breast-ovarian cancer syndrome,
  
• site-specific ovarian cancer syndrome, and
  
• hereditary nonpolyposis colorectal cancer syndrome (HNPCC) (Lynch II).²
  

The first 2 are caused by inherited mutations in the BRCA1 and BRCA2 genes. In fact, though often described as separate entities, these syndromes are likely phenotypic variants of the same genetic mutations. BRCA1 and BRCA2 function as classic tumor-suppressor genes and are inherited in an autosomal dominant fashion.

HNPCC is caused by mutations in a series of genes responsible for repairing errors in DNA replication. Inactivation of these so-called mismatch repair genes results in a high incidence of right-sided colon cancer, endometrial cancer, and ovarian cancer.³

Gene mutations in different populations

The lifetime risk of developing ovarian cancer in the United States is about 1.4%. However, among women with BRCA1 or BRCA2 mutations, the risk rises to 20% to 60%.⁴

These genes also impart a significant lifetime risk of breast cancer in women and, in the case of BRCA2, in men as well.

Less than 0.15% of the general population carries BRCA1 or BRCA2 mutations. However, the carrier rate is dependent on ethnic background.⁵ Founder mutations have been identified among multiple unrelated families in Iceland, the Netherlands, and Sweden, and among Jews of Central or Eastern European descent (Ashkenazi).

The best described founder mutations are the 185delAG and 5382insC mutations in BRCA1 and the 6174delT mutation in BRCA2, occurring in Ashkenazi Jews at a carrier rate of 2%.⁶

Ovarian cancer in Ashkenazi women more likely genetic than sporadic

Although an Ashkenazi woman is no more likely to develop ovarian cancer than a noncarrier, if she does develop the disease, it is far more likely to be genetic rather than sporadic.

Consequently, if a woman of Ashkenazi Jewish descent develops ovarian cancer, there is a 40% chance she carries a mutation in one of these 2 genes.⁷ Her first-degree relatives (mother, sisters, daughters) have a 20% risk of being gene carriers (50% in autosomal dominant transmission).

Therefore, an Ashkenazi Jewish woman needs only 1 first-degree relative with ovarian cancer to be considered for further genetic counseling.

 

The 3-2-1 rule for genetic counseling

Genetic counseling is indicated when the patient meets the “3-2-1” rule, known as the modified Amsterdam criteria:

• 3 affected individuals with either colorectal or ovarian cancer, in
  
• 2 successive generations, with at least
  
• 1 who developed cancer under the age of 50 years.
  

An Ashkenazi woman with only 1 first-degree relative with ovarian cancer is eligible for genetic counseling.

When HNPCC is present

Women with documented HNPCC have a 70% lifetime risk of developing endometrial cancer and a lifetime risk of ovarian cancer of 11% or more. Testing for mutation in mismatch repair genes can be performed on peripheral leukocytes. Alternatively, the primary tumor from affected individuals can be assessed for the presence of microsatellite instability, a consequence of defective mismatch repair.³

How to obtain a family history

Women with a significant family history of breast, colorectal, endometrial, or ovarian cancer may face an elevated risk of ovarian cancer. Basic questions about the occurrence of these malignancies in first-degree (mother, sister, daughter) or second-degree (aunt, grandmother) relatives must be a component of every gynecologic history.

1-question assessment

The following question is a simple but effective way to assess familial cancer risk: “Has anyone in your family had breast cancer under the age of 35, or colorectal, uterine, or ovarian cancer?” If so, ask further questions about the relative’s age at onset and how the affected person is related to the patient.

Although an informal pedigree can be constructed in a few moments, a formal pedigree is a more daunting task. More than half of family histories of ovarian cancer are inaccurate—and the error rate increases if the affected family member is a distant relative.⁸

Moreover, not all ovarian cancers are associated with genetic syndromes. Serous epithelial ovarian, fallopian tube, and peritoneal cancers dominate in BRCA-associated syndromes, whereas mucinous epithelial cancers are very rare. Germ-cell and stromal neoplasms as well as epithelial tumors of low malignant potential are excluded when assessing genetic risk.

Additional documentation is often necessary to verify the history. Such medical detective work may be beyond what a generalist is willing to perform. In that case, referral to a genetic counselor is the next step.

When to refer

Recommendation. If she is of Ashkenazi Jewish descent, like the patient described at the beginning of this article, only 1 first-degree relative indicates referral for genetic counseling.

If the patient claims at least 2 first-degree or second-degree relatives with breast cancer under the age of 35, or with ovarian cancer, suspect a genetic syndrome.

Family histories of particular concern include breast and ovarian cancer in a single individual, or any case of male breast cancer. Refer these patients to genetic counseling for formal pedigree analysis and possible testing for genetic predisposition. In the case of HNPCC, refer the patient if the modified Amsterdam criteria are met.

What does a genetic counselor do?

One of the most important roles of a genetic counselor is constructing as accurate a pedigree as possible. In some cases, this requires much effort on the part of both patient and counselor. It is not uncommon for the counselor to request outside pathology reports, operative notes, and death certificates to verify cancer cases. Pathology slides also may be necessary, while tissue blocks occasionally are requested to assist in genetic testing.

When testing is warranted

Genetic predisposition testing should be undertaken only after formal consultation with a genetic counselor who has expertise in cancer predisposition syndromes. As a policy statement from the American Society of Clinical Oncology concludes, genetic testing for cancer susceptibility should be performed only when:

Obesity is associated with increased risk of ovarian cancer mortality.

• pre- and post-test counseling is available,
  
• the test can be interpreted, and
  
• the results will help in medical and surgical management.
  

Ideally, testing should take place in the setting of a multidisciplinary team with expertise in the interpretation of verified family cancer pedigrees and in the medical, emotional, financial, and legal ramifications of genetic testing.⁹

CASE 2

A 65-year-old obese nullipara is currently on year 10 of combined hormone replacement therapy (HRT). Although she has no family history of breast or ovarian cancer, she asks about her risk of developing ovarian cancer.

Other risk factors

Findings on the following potential risk factors for ovarian cancer may help address this patient’s concerns:

Dietary factors

In a recent study, obesity was associated with an increased risk of ovarian cancer mortality.¹⁰ Women who eat a diet high in saturated fat and low in vegetable fiber also may face an increased risk.

 

In 1989, the observation that Swedish women had both a high risk of ovarian cancer and the highest per capita dairy consumption in the world led some investigators to postulate a relationship between lactose consumption and ovarian cancer. The reason: When compared with matched controls, women with ovarian cancer were more likely to have high levels of galactose, a component sugar of the disaccharide lactose and a known oocyte toxin.¹¹

This observation, however, has been inconsistent. Therefore, no specific dietary strategy can be recommended to reduce the risk of ovarian cancer.

Recommendation. I would advise this patient to maintain a normal body mass index. Data are insufficient to support more specific diet recommendations.

Talc exposure

When talc is placed on the perineum, it may enter the vagina and ascend to the upper genital tract. Because talc is structurally similar to asbestos, it may theoretically increase the ovarian cancer risk. The observation that women who undergo tubal sterilization procedures or hysterectomy have a lower risk of ovarian cancer supports the ascending-carcinogen hypothesis.

Multiple case-control studies have shown a small but consistent increased risk with perineal application of talc (odds ratio 1.3, 95% confidence interval [CI], 1.1–1.6).¹² The risk appears to be time- and dose-dependent, with greater risk associated with more frequent application of perineal talc over a long duration.

Recommendation. The practice of applying genuine talc to the perineum should be discouraged. Cornstarch-based dusting powders are widely available.

Infertility drugs

One of the most difficult issues to study is the relationship between infertility drugs and ovarian cancer, although we know that unexplained infertility is an independent risk factor for ovarian cancer.

A retrospective study that claimed an association between prolonged clomiphene exposure and ovarian cancer¹³ was not restricted to invasive epithelial ovarian cancers, but included granulosa cell tumors. These estrogen-secreting neoplasms of stromal origin may directly contribute to infertility by disrupting normal follicular maturation and the menstrual cycle.

A number of studies—including a large collaborative analysis of 12 case-control studies—have reported an association between fertility drugs and invasive epithelial ovarian cancer.¹⁴ In addition, many of the theoretical models of epithelial ovarian cancer pathogenesis implicate both incessant ovulation and high gonadotropin levels as important steps in malignant transformation of ovarian epithelium.

Oral contraceptives (OCs), which reduce ovulatory events, and moderate gonadotropin levels are associated with a consistent and significant protective effect.

Recommendation. It seems prudent, in the absence of convincing data, to use fertility medication only when absolutely indicated, at the lowest effective dose, and for the shortest duration possible without compromising treatment success.

However, prior exposure to these agents should not be considered an indication for increased surveillance or prophylactic surgery.

Estrogen replacement therapy

Women on estrogen replacement therapy appear to have an increased risk of ovarian cancer. When compared to nonusers, “ever-users” had a relative risk of ovarian cancer of 2.² (95% CI, 1.53–3.17), and the risk increased with the duration of use.¹⁵ Long-term users, defined as women who used estrogen replacement therapy for at least 20 years, had a relative risk of 3.2 (95% CI, 1.7–5.7).¹⁶

Although some studies suggest a protective effect of combination hormone replacement regimens that include both estrogen and progesterone, this observation has not been confirmed. Thus, long-term estrogen users should consider an increased risk of developing ovarian cancer when deciding whether to initiate or continue estrogen replacement therapy.

Recommendation. I would advise the Case 2 patient to stop HRT.

Primary prevention

Oral contraceptives

OCs reduce the risk of ovarian cancer significantly. A number of studies have demonstrated a 10% risk reduction per year for up to 5 to 7 years of use.¹⁷ This effect appears to persist for at least 10 years after OCs are discontinued. It also has been observed in patients known to be BRCA1 and BRCA2 carriers and is the basis for recommending OCs as a chemoprophylactic method in known carriers who wish to retain fertility.¹⁸

Long-term estrogen users should consider increased risk of ovarian cancer. Ovarian cancer: High-risk patients.

However, use of OCs by BRCA carriers is not without some controversy. An Israeli population-based study¹⁹ of ovarian cancer and OC use demonstrated a protective effect of pregnancy but not use of OCs. It is unclear why the Israeli data is inconsistent with prior published reports.

Recommendation. In Case 1, the patient has taken OCs for 9 years. Her risk has been substantially reduced, and I would advise her to continue OC pills when not attempting to get pregnant, until she is near menopause.

Prophylactic surgery

In high-risk women, preventive surgery may substantially reduce but not completely eliminate the risk of ovarian cancer. For example, bilateral salpingo-oophorectomy in BRCA carriers reduces the risk of ovarian cancer by more than 90% and the risk of breast cancer by more than 50%.^20,21

 

The operation should be reserved for women with known mutations in BRCA1 or BRCA2 or who have a family history consistent with one of the genetic syndromes associated with ovarian cancer.

The addition of hysterectomy does not appear to increase the efficacy of the operation and should be performed only for concurrent gynecologic indications or if the patient has HNPCC.

Patients should be informed that prophylactic surgery does not protect against subsequent papillary serous carcinoma of the peritoneum. They also should be warned that about 7% of operations detect occult ovarian or tubal carcinoma, which may not be identified until final pathology reports are issued.²² Pathologists should be instructed to submit the entire specimen for sectioning to reduce risk of missing microscopic occult malignancy.

In addition, the patient should be prepared for surgical menopause.

Effective primary prevention strategies such as chemoprophylaxis and prophylactic surgery, when appropriately applied, may spare many women the devastating consequences of this dreaded disease.

Internet resources

• Ovarian Cancer (PDQ): Prevention of ovarian cancer http://www.cancer.gov/cancerinfo/pdq/prevention/ovarian/healthprofessional
  
• Myriad Genetics: BRCA1/2 mutation prevalence tables http:///www.myriadtests.com/provider/mutprevo.htm
  
• National Society of Genetic Counselors: Locate a genetic counselor specializing in cancer risk assessment http://www.nsgc.org/resourcelink.asp
  

The author reports no financial relationships relevant to this article.

The outlook for preterm premature rupture of membranes (PPROM) has improved considerably since a landmark study showed clear benefits of antibiotics.

Previously, approximately 80% of women with PPROM experienced spontaneous labor within 48 hours with expectant management.

We now know that infection is a major cause of PPROM and that antibiotic therapy decreases neonatal morbidity and increases the interval from rupture of membranes to delivery. These benefits are most evident at early gestational ages.

This article reviews newer studies, as well as that breakthrough 1997 study, and their implications for diagnosis and treatment, including optimal drug regimens. Recommendations for 4 special situations are also discussed:

• presence of cerclage,
• herpes simplex infection,
• bleeding, and
• outpatient management.

Despite progress, research is still needed. For example, because much of our information about clinical interventions for PPROM, such as corticosteroid therapy and use of tocolytics, predates the use of antibiotics, many earlier studies need to be repeated.

A variety of methods confirm the diagnosis

Approximately 90% of patients with PPROM report nonurinary fluid leaking from the vagina.^1,2 Nevertheless, a history of leaking fluid should be confirmed by examination.

Nitrazine paper test

The most common way to diagnose rupture of membranes involves exposing nitrazine paper to the leaking fluid. If the fluid is alkaline, the paper will turn bright blue.

However, seminal fluid, urine, and blood can also turn nitrazine paper blue. Therefore, a confirmatory method should be used with nitrazine paper testing.

Fern pattern test

When amniotic fluid proteins are allowed to dry and then examined under a microscope, they will exhibit a “fern” pattern.

The combination of nitrazine paper and the fern test has a specificity of over 90%.

Avoid digital cervical examination

Digital examination during testing may diminish latency (the period from rupture of membranes to delivery) and increase neonatal morbidity.³

Testing at early gestational ages

Occasionally, PPROM occurs early in the middle trimester. At such early gestations, amniotic fluid may not fern or produce the classic blue color when exposed to nitrazine paper.

Other tests may be used, however:

• Ultrasound can help in evaluating the quantity of amniotic fluid.
• Indigo carmine dye can be placed in the amniotic cavity, with a tampon inserted in the vagina. If the tampon turns blue when the patient ambulates, rupture of membranes is confirmed.
• Alpha-fetoprotein and human choriogonadotropin. When alpha-fetoprotein is present and human choriogonadotropin is highly concentrated in amniotic fluid, ruptured membranes are confirmed.⁴

Fortunately, a large majority of patients can be diagnosed using clinical history, nitrazine paper, and the fern pattern test.

 

Gestational age determines management

The number 1 factor determining management is gestational age.

Before 23 weeks, the pregnancy is considered previable

Counsel these patients about the risk of infection, the prognosis of extreme prematurity, and the generally poor outcome of these pregnancies, which can involve pulmonary hypoplasia secondary to severe oligohydramnios (FIGURE 1).

Induction of labor is an option to terminate the pregnancy. If the patient chooses expectant management, monitor her closely for infection.⁴ Outpatient management is acceptable for the highly compliant patient. However, once fetal viability is attained, readmit the patient to monitor maternal and fetal well-being.

Between 23 and 32 weeks

Treat these women expectantly in the hospital unless there is evidence of infection or fetal compromise. Management includes intravenous antibiotic therapy and corticosteroids, as recommended by the National Institutes of Health (FIGURE 2).⁵

Although it is controversial, consider magnesium sulfate tocolytic therapy if the patient is actively contracting with no evidence of infection.⁶

Some authorities recommend amniocentesis (at any gestational age) to rule out infection prior to expectant management, while others consider the appropriate use of clinical parameters to be adequate.⁶

After monitoring the patient in the labor suite for 24 to 48 hours, transfer her to the antenatal ward, if she is stable, for frequent fetal surveillance and maternal evaluation. Continue to evaluate her for signs of intraamniotic infection such as fever, uterine tenderness, and foulsmelling discharge.

Many authorities recommend daily nonstress testing or biophysical profiles to evaluate fetal well-being.⁷

Between 32 and 34 weeks

Management for this gestational age range is the most controversial (FIGURE 3). Mercer et al⁸ demonstrated that expectant management is harmful to the neonate if fetal lung maturity is present. Most experts contend that some method of fetal lung assessment should be undertaken. The most effective method is sampling amniotic fluid by amniocentesis. Not only can fetal lung maturity be determined, but the fluid can be analyzed for evidence of infection using the Gram stain, glucose level, and white blood cell count.

When an amniocentesis site is unavailable, consider assessing fetal lung maturity via a pooled amniotic fluid sample from the vagina.⁹ Deliver the infant if either intraamniotic infection or fetal lung maturity is present.

If there is no evidence of fetal lung maturity, expectant management is usually recommended for pregnancies between 32 and 34 weeks.

In a recent survey of maternal-fetal medicine specialists in the United States, 58% recommended delivery at 34 weeks of gestation.¹⁰

For patients who are at less than 34 weeks upon presentation—or to be treated expectantly—consider giving antibiotics, corticosteroids, and tocolysis during the first 48 hours.

FIGURE 1 Managing premature rupture of membranes: Less than 23 weeks’ gestation

FIGURE 2 Managing premature rupture of membranes: 23 to <32 weeks’ gestation

The new standard of care: antibiotic therapy

In the 1980s, several studies suggested antibiotics were beneficial for PPROM patients.¹¹ More recently, the Maternal-Fetal Medicine Network published the results of a large, prospective, randomized, double-blind, placebo-controlled trial clearly demonstrating that antibiotics prolong the latency period and decrease neonatal morbidities (see “Antibiotics in PPROM improve outcomes in Maternal-Fetal Medicine Network study”).¹²

Limited- versus broad-spectrum antibiotics

Although antibiotic therapy is now the standard of care for patients with PPROM, it remains unclear which antibiotic is best. Some researchers have recommended limited-spectrum antibiotics, while others prefer a broader-spectrum approach.¹³ Several recent studies¹⁴ have suggested that erythromycin is effective for eradicating Mycoplasma and Ureaplasma and beneficial for the neonate.

Optimal duration of antibiotic therapy is unclear

Most early studies advocated standard treatment from the time of ruptured membranes until delivery, while others recommended 7 to 10 days of therapy. Several recent trials¹⁵ have found 3 days of therapy to be equivalent to 7 days. However, these studies lack sufficient power to warrant a shorter duration of therapy at this time.

 

Corticosteroids beneficial if infection is treated

In 1998, the National Institutes of Health issued a Consensus Statement⁵ recommending corticosteroid therapy to accelerate fetal lung maturity and decrease neonatal morbidity in PPROM patients. Many authorities questioned the wisdom of this recommendation, since earlier studies had suggested an increased risk of maternal morbidity when corticosteroids are given as therapy for PPROM.¹⁶ However, these early studies did not use antibiotics to treat infection—the primary cause of PPROM—and many patients underwent digital vaginal exams, which probably contributed to unfavorable maternal and neonatal outcomes.

I recently began treating PPROM patients with antibiotics for 12 hours prior to administering corticosteroids and noted a considerable benefit for steroid therapy.¹⁷ I therefore recommend that PPROM patients who are treated expectantly receive corticosteroids unless there is evidence of an infection.

Tocolysis: jury still out

The use of tocolytics also is controversial. Many original studies reported no overall benefit and a significant increase in maternal morbidity when tocolytic therapy was given for ruptured membranes.¹⁸ However, many patients in these trials underwent digital vaginal exams and were not given antibiotics, and many received corticosteroids.

In my practice, I give magnesium sulfate and corticosteroids for the first 48 hours to women at or beyond 23 weeks’ gestation, provided there is no evidence of infection.¹⁹ However, additional data is needed before this management protocol can be considered the standard of care.

Assess fetal well-being frequently

Initial recommendations for antepartum testing were based on the high perinatal morbidity and mortality associated with PPROM. Nonstress testing was the foundation of therapy, and many authorities advocated daily nonstress testing even in the absence of definitive proof of its benefit.⁴

In the late 1980s, Vintzileos²⁰ and other experts recommended daily biophysical profiles, suggesting that this strategy could identify patients with subclinical infection.

Daily biophysical profiles: No benefit beyond that of daily nonstress test

My colleagues and I⁷ conducted the only reported prospective, randomized trial comparing daily biophysical profiles to daily nonstress testing with backup biophysical profiles for abnormalities. We concluded that daily biophysical profiles provide no benefit beyond that achieved with daily nonstress testing. Whether a less intense testing protocol would produce the same benefit is unclear.

Important to recognize high risk of poor perinatal outcome

The important point in managing patients with PPROM is recognizing the high risk of poor perinatal outcome. For this reason, frequent evaluation of fetal well-being is an essential element of any management plan.

Determine mode of delivery as usual, but monitor fetus

In PPROM cases, the fetus often is intolerant of labor, primarily because of oligohydramnios and subclinical infection. For this reason, monitor the fetus very closely in the intrapartum period. Amnioinfusion decreases variable decelerations and improves pH in these patients and should be considered when it is clinically indicated.²¹

Determine the mode of delivery using routine obstetrical indications. Deliver viable breech infants by cesarean section.

Few data exist on the benefit of cesarean for the very premature neonate; use of routine obstetrical indicators is advisable.

Special situations

Four conditions may affect management of the patient with PPROM:

Cerclage in place

Numerous early studies suggested that this foreign body is a focus of infection and recommended removal. However, recent reports have not substantiated these findings.²² A large, multicenter, prospective, randomized trial is underway, which compares expectant management with cerclage removal when membranes rupture. It should provide a definitive answer.

Herpes simplex infection

If the patient with PPROM has a clinical outbreak of herpes simplex virus, weigh the risk of early delivery against the risk of herpes simplex infection. Major et al²³ compiled a case series showing that infants did not become infected with herpes simplex virus when women were managed expectantly. Still, consider prophylactic therapy with antiviral agents in expectantly managed patients. At later gestational ages, most experts recommend delivery.

Bleeding

Placental abruptions occur in about 5% of PPROM pregnancies.²⁴ Some authorities contend that the cause of abruption is placental shearing following leakage of amniotic fluid and decreased intrauterine volume. However, bleeding is rarely substantial enough to warrant delivery. Even so, monitor women with active bleeding in the labor suite, remaining vigilant for evidence of fetal compromise.

Outpatient management

Only 1 study has evaluated outpatient management of women with PPROM.²⁵ Until more definitive information is available, the American College of Obstetricians and Gynecologists recommends that outpatient management be limited to approved study protocols.⁶

The author reports no financial relationships relevant to this article.