Cleveland Clinic Journal of Medicine

As the nation gets heavier, our livers will get fattier. The prevalence of nonalcoholic fatty liver disease (NAFLD) has been rising in tandem with the rise in obesity ever since the term nonalcoholic steatohepatitis (NASH, a subtype of NAFLD) was coined by Ludwig in 1980.¹ Yet, despite an explosion of research on NAFLD and gains in understanding its epidemiology and pathogenesis, a number of issues remain unresolved, including how to treat it.

NAFLD IS A SPECTRUM

NAFLD is a spectrum. The mildest form is simple fatty liver, or simple steatosis. Next is NASH, or fatty liver with inflammation and evidence of damage to hepatocytes (liver cells). Still more severe is cirrhosis, and in its most extreme form NAFLD can progress to hepatocellular carcinoma or liver failure. The distinction between simple steatosis and NASH is important because their prognoses and management are different.

NAFLD IS COMMON AND LINKED TO OBESITY

NAFLD is the most common cause of elevated liver enzymes and also one of the most common forms of liver disease in the world. It is now estimated to affect about 20% to 30% of people in the United States and other Western countries. In contrast, the prevalence of chronic hepatitis C virus infection is estimated at 3% of the world’s population. In comparison to the prevalence of NAFLD, the prevalence of NASH is much lower: 2% to 3% in the United States.² The incidence of NAFLD is expected to rise further with the increase in obesity in the United States.

NAFLD is even more common in people who are morbidly obese, ie, who have a body mass index greater than 40 kg/m². In a series of studies of morbidly obese patients undergoing bariatric surgery (N = 1,620), the prevalence of hepatic steatosis was 91% (range 85%–98%), and the prevalence of NASH was 37% (range 24%–98%). NASH was not predicted by age or body mass index, but it was more common in men, people with diabetes, and people with insulin resistance.³

Obesity is also increasing in prevalence in children. Since liver biopsies were not done in most pediatric studies, the pediatric prevalence data are based on elevated aminotransferase levels and on ultrasonographic findings of echogenic livers. The overall prevalence of NAFLD in children is estimated at 3% to 10%, but it may be much higher in obese children.⁴

Arun et al⁵ found that the prevalence of NASH in morbidly obese men was almost twice as high as in morbidly obese women (60.3% vs 30.9%). In contrast, earlier studies suggested that NAFLD was more prevalent in women. This higher incidence of NASH may also reflect the higher incidence of metabolic syndrome in morbidly obese men (91.4% vs 76.2%).

Less common in African Americans

In the United States, African Americans have consistently been found to have the lowest prevalence of NAFLD. In a California population study of 159 newly diagnosed NAFLD cases, non-Hispanic whites accounted for 45%, followed by Hispanics (28%), Asians (18%), and African Americans (3%). After controlling for the ethnic composition of the entire cohort, Hispanics had the highest rate of NAFLD and African Americans the lowest.⁶ In Eastern countries such as Japan, the prevalence of NAFLD is estimated to be about 9.3%. Interestingly, about half of the people with NAFLD in Japan were not overweight.⁷

The difference in prevalence of NAFLD in different ethnic groups may be explained by their different rates of metabolic syndrome (21.6% in African Americans vs 23.8% in whites vs 31.9% in Mexican Americans⁸) as well as other genetic and environmental factors.

NAFLD IS USUALLY CLINICALLY SILENT

NAFLD is usually clinically silent, and its impact has most likely been underestimated. Symptoms, if present, are minimal and non-specific, such as fatigue and right upper quadrant discomfort. Most findings on physical examination are also normal. Most patients seek care because of an incidental finding of elevated aminotransferase levels or radiographic studies suggesting the liver is fatty.⁹

The estimated prevalence of aminotransferase elevations in the general population from the third National Health and Nutrition Examination Survey data is 7.9%,¹⁰ with about two-thirds of cases unexplained. Of the unexplained cases, most are strongly associated with metabolic syndrome and probably represent underlying NAFLD.¹⁰

Yet aminotransferase levels are typically normal or elevated by less than five times the upper limit of normal (usually < 250 IU/L).⁹ In contrast to those with alcoholic hepatitis, most patients with NAFLD have a ratio of aspartate aminotransferase to alanine aminotransferase of less than 1. As the disease progresses, the aspartate aminotransferase level increases more than the alanine aminotransferase level, so if the ratio is more than 1, more advanced liver disease may be suspected.¹¹

Levels of other liver enzymes such as alkaline phosphatase and of acute-phase reactants such as ferritin may also be elevated. Ferritin is believed to reflect hepatic injury, inflammation, or insulin resistance.

A DIAGNOSIS OF EXCLUSION

Excessive alcohol consumption must especially be excluded. Most studies defined excessive alcohol consumption as more than 20 to 40 g/day.² Recently, this threshold has been lowered to 20 g/day (roughly two drinks) in men and 10 g/day in women.

Insulin resistance should be estimated, given the close relationship between NAFLD and insulin resistance and the metabolic syndrome. Insulin resistance can be measured accurately in a number of ways. The Homeostasis Model Assessment is an easy method that provides an estimate of insulin resistance based on fasting serum glucose and serum insulin levels.¹³

Serologic tests can rule out hepatitis B and hepatitis C. In those with negative results, especially in those with components of the metabolic syndrome or insulin resistance, NAFLD is responsible for most cases of persistently elevated serum liver enzymes.

Imaging tests

Radiographic evaluation is another noninvasive way to diagnose fatty liver. The sensitivity of either ultrasonography or computed tomography for detecting hepatic steatosis is between 93% and 100% when there is more than 33% fat in the hepatic parenchyma.¹⁴ None of the radiographic methods, including magnetic resonance imaging, can accurately differentiate between nonprogressive simple steatosis and NASH, but the technology is advancing. Contrast ultrasonography and magnetic resonance spectroscopy have shown promise and may become useful in the future.

Other noninvasive tests

Ultrasonographic elastrography (FibroScan), a noninvasive way to measure liver stiffness, has also been used in patients with hepatitis C. Although the preliminary data in NAFLD are interesting, additional validation is needed.

Serum biomarkers, including markers of fibrosis (eg, FibroSURE), apoptosis, and adipocytokines have been used to diagnose NASH. The markers of apoptosis are especially interesting but need further validation.

Liver biopsy remains the gold standard

Because we lack a fully validated noninvasive biomarker of NASH, liver biopsy remains the gold standard for diagnosing it. The minimum histologic criteria for establishing the diagnosis of NASH have been debated; most pathologists require at least 5% hepatic steatosis, mixed lobular inflammation, and hepatocellular ballooning.

In a study of 354 liver biopsies of patients with negative results on serologic tests, NASH was found in 34% and fatty liver in 32%. In the same study, the findings on liver biopsy led to alterations in patient management in 18% of cases.¹⁵

Some clinicians doubt the value of liver biopsy in patients with suspected NASH, in view of possible sampling error in the biopsy specimens (the distribution can be patchy, and if the specimen is taken from an unaffected area, the results can be falsely negative) and because there is no established effective therapy for NAFLD. However, liver biopsy is the only test that can accurately establish the diagnosis of NASH and tell us the stage of liver disease, which has important prognostic implications. Most experts agree that liver biopsy should be considered for patients at risk of advanced liver disease, such as those with persistently elevated liver enzyme levels despite intervention to reverse conditions associated with metabolic syndrome.¹⁶

PATHOGENESIS: THE MULTIPLE-HIT HYPOTHESIS

NAFLD is closely linked to obesity, insulin resistance, and metabolic syndrome.¹³ Insulin allows free fatty esterification and triglyceride fat storage in adipose tissues. When insulin resistance develops, free fatty acids are inappropriately shifted to nonadipose tissues, including the liver. Insulin resistance increases free fatty acid flux to the liver by decreased inhibition of lipolysis and also increased de novo lipogenesis.¹⁷

Insulin resistance and visceral obesity also result in decreased levels of a “protective adipokine,” adiponectin. Adiponectin inhibits liver gluconeogenesis and suppresses lipogenesis. Thus, decreased adiponectin hinders fatty acid oxidation and increases fat accumulation in the liver. Other adipocytokines that are important in NAFLD are resistin, leptin, visfatin, tumor necrosis factor alpha, and interleukin 6.

Apoptosis and oxidative stress may also contribute to the development and progression of NASH. In this context, the “multiple-hit hypothesis” for the pathogenesis of NASH has become quite popular.¹⁸ An in-depth review of the pathogenesis of NAFLD is beyond the scope of this paper; readers are referred to a recently published review on this subject.¹⁹

STEATOSIS IS BENIGN, BUT NASH CAN PROGRESS

Simple steatosis by itself generally has a benign prognosis. In a 1995 cohort study with a median follow-up of 11 years, there was no progression of simple steatosis to NASH or cirrhosis,²⁰ and recent reviews estimate that only a small portion of patients with simple steatosis develop steatohepatitis. The validity of these data is still being debated.

On the other hand, once patients have progressed to NASH, histologic progression has been noted in about 32% to 41% of patients over a median follow-up of 4.3 to 13.7 years.^21,22 This would mean that approximately 9% of patients with NASH may develop cirrhosis.²¹

People with cirrhosis due to NAFLD are at risk of developing liver-related morbidity and of death. In one of the longest follow-up cohort studies (mean follow-up of 13.7 years), end-stage liver disease developed in 5.4%, and hepatocellular carcinoma developed in about 2%. About 20% of the patients died, with more than 70% of the deaths in patients who had NASH at baseline. The survival rate was lower in patients with NASH, whereas no difference in survival was seen in the group with simple steatosis.²²

A number of studies have assessed independent predictors of advanced fibrosis. Most studies suggest that elevated liver enzymes, metabolic syndrome, or type 2 diabetes is associated with advanced liver disease. Although noninvasive biomarkers of fibrosis have been developed for hepatitis C, to date, a fully validated, noninvasive biomarker of fibrosis for NAFLD does not exist.

As noted, the spectrum of NAFLD also includes hepatocellular carcinoma, and in a series of 105 patients with hepatocellular carcinoma, hepatitis C virus accounted for 51% and cryptogenic liver disease accounted for another 29%. Since cases of cryptogenic cirrhosis in the United States are considered to be “burned out NASH,” approximately 13% of patients with hepatocellular carcinoma may have had underlying NAFLD as the cause of their liver disease.²³ These data suggest that, similar to other cirrhotic patients, NAFLD patients with cirrhosis should be screened for hepatocellular carcinoma.

NO CONSENSUS ON TREATMENT

Weight loss

Modest weight loss—less than 2 pounds (1 kg) per week—is associated with a decrease in the incidence of metabolic syndrome and can also improve the histologic features of NASH in more than 80% of cases.²⁴ Loss of as little as 4% to 5% of body weight is also associated with lowering of aminotransferase and fasting insulin levels.²⁵

The mechanism of benefit is via loss of adipose tissue, which decreases insulin resistance. Weight loss by any means, including bariatric surgery for morbid obesity or use of weight-reducing agents, has been correlated with improvement in liver enzyme levels, liver histologic findings, or both.^24,26

However, the traditional low-calorie, low-fat diet may not be optimal for NAFLD patients. In one study,²⁷ patients consuming more than 54% of their calories from carbohydrates compared with those consuming less than 35% had an odds ratio of 6.5 for hepatic inflammation. This finding is not surprising in light of prior research in which high carbohydrate intake increased hepatic de novo lipogenesis. On the other hand, there was no association between total caloric or protein intake and hepatic steatosis or fibrosis. Contrary to traditional beliefs, patients with higher fat intake had less inflammation, steatosis, and fibrosis.

Insulin sensitizers

Given that insulin resistance seems to be the main pathophysiologic culprit in NAFLD, two classes of insulin sensitizers have been studied:

Biguanides act mainly by increasing hepatic insulin sensitivity and reversing insulin resistance induced by tumor necrosis factor alpha.

Glitazones improve insulin sensitivity in both diabetic and euglycemic patients by activating the nuclear transcription factor called peroxisome proliferator-activated receptor (PPAR) gamma.

Both biguanides and glitazones have been found to lower liver enzyme levels, decrease insulin resistance, and improve histopathologic findings. However, the effects of glitazones do not persist after the drugs are stopped, and these drugs and are also associated with an average weight gain of 3 to 6 kg.^28,29

Although these data are encouraging, they are preliminary, and more evidence is needed to establish the safety and efficacy of these drugs in treating patients with NASH.

Antioxidants

Antioxidants such as vitamin E, n-acetyl-l-cysteine, s-adenosylmethionine (SAMe), and betaine have been investigated in the treatment of NAFLD.

Vitamin E has been most widely studied. Being fat-soluble, vitamin E can stabilize mitochondrial function and is theorized to inhibit lipid peroxidation and subsequent free radical reactions. Smaller, nonrandomized trials have found that vitamin E improves biochemical markers of liver inflammation. However, in one of the largest randomized controlled trials (with 45 patients), patients taking vitamin E showed improvement in their fibrosis scores but no differences in their necroinflammatory activity or alanine aminotransferase levels.³⁰ Most studies of antioxidants show at least mild improvement in biochemical or histologic signs of NAFLD.³¹

SAMe and betaine are important antioxidants. However, most studies of SAMe and betaine have been small and inconclusive.

Two large phase III clinical trials are under way at the National Institute of Diabetes and Digestive and Kidney Diseases. They should clarify the role of these agents in the treatment of NASH. The PIVENS (Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients With Nonalcoholic Steatohepatitis) study has completed enrollment of 240 patients, but the final data are not available. The second study, TONIC (Treatment of Nonalcoholic Fatty Liver Disease in Children) will be one of the largest studies of NAFLD in children; it will be looking at vitamin E, metformin, or placebo over a 2-year follow-up. The TONIC study is still under way, so the final data are not yet available.

Ursodeoxycholic acid, another cytoprotective agent, has traditionally been used for primary biliary cirrhosis, but the data are conflicting on its efficacy in NAFLD. Of note, some bile acids are hepatotoxic and facilitate apoptosis via a Fas ligand-mediated pathway. On the other hand, ursodeoxycholic acid is a hydrophilic bile acid that may act to displace the hepatotoxic hydrophobic endogenous bile acids and potentially has an antiapoptotic and cytoprotective effect in NAFLD. Although liver enzyme levels declined in a few of the studies of ursodeoxycholic acid in patients with NAFLD, a large randomized clinical trial (in 166 patients) did not show any significant difference from placebo in liver enzyme levels or liver histologic findings.³²

Lipid-lowering drugs

Lipid-lowering drugs target the high levels of triglycerides and low levels of high-density lipoprotein cholesterol that often occur in insulin resistance and metabolic syndrome associated with NAFLD. A few small studies found that aminotransferase levels fell with both statins and gemfibrozil (Lopid).³³ Even if liver enzyme levels are abnormal, most experts believe that statins are relatively safe to use in patients with NAFLD who need cholesterol-lowering agents. Nevertheless, clinical monitoring of these patients for potential hepatic toxicity is recommended.

Other medications

Other medications, such as pentoxifylline (Pentoxil, Trental), probiotics, and angiotensin-converting enzyme inhibitors, have been used in small studies of patients with NASH, with encouraging but inconclusive results.

Although a number of pilot studies of agents for treating NAFLD have been proposed, they are small and open-label. With the tremendous recent gains in clinical investigation, functional genomics, and proteomics, it is expected that our understanding of NASH and its treatment will be broadened.

In summary, despite the relatively large number of agents tested for the treatment of NAFLD, most of the data are preliminary. Thus, in 2008, there is no established, evidence-based treatment for patients with NASH.

As the nation gets heavier, our livers will get fattier. The prevalence of nonalcoholic fatty liver disease (NAFLD) has been rising in tandem with the rise in obesity ever since the term nonalcoholic steatohepatitis (NASH, a subtype of NAFLD) was coined by Ludwig in 1980.¹ Yet, despite an explosion of research on NAFLD and gains in understanding its epidemiology and pathogenesis, a number of issues remain unresolved, including how to treat it.

NAFLD IS A SPECTRUM

NAFLD is a spectrum. The mildest form is simple fatty liver, or simple steatosis. Next is NASH, or fatty liver with inflammation and evidence of damage to hepatocytes (liver cells). Still more severe is cirrhosis, and in its most extreme form NAFLD can progress to hepatocellular carcinoma or liver failure. The distinction between simple steatosis and NASH is important because their prognoses and management are different.

NAFLD IS COMMON AND LINKED TO OBESITY

NAFLD is the most common cause of elevated liver enzymes and also one of the most common forms of liver disease in the world. It is now estimated to affect about 20% to 30% of people in the United States and other Western countries. In contrast, the prevalence of chronic hepatitis C virus infection is estimated at 3% of the world’s population. In comparison to the prevalence of NAFLD, the prevalence of NASH is much lower: 2% to 3% in the United States.² The incidence of NAFLD is expected to rise further with the increase in obesity in the United States.

NAFLD is even more common in people who are morbidly obese, ie, who have a body mass index greater than 40 kg/m². In a series of studies of morbidly obese patients undergoing bariatric surgery (N = 1,620), the prevalence of hepatic steatosis was 91% (range 85%–98%), and the prevalence of NASH was 37% (range 24%–98%). NASH was not predicted by age or body mass index, but it was more common in men, people with diabetes, and people with insulin resistance.³

Obesity is also increasing in prevalence in children. Since liver biopsies were not done in most pediatric studies, the pediatric prevalence data are based on elevated aminotransferase levels and on ultrasonographic findings of echogenic livers. The overall prevalence of NAFLD in children is estimated at 3% to 10%, but it may be much higher in obese children.⁴

Arun et al⁵ found that the prevalence of NASH in morbidly obese men was almost twice as high as in morbidly obese women (60.3% vs 30.9%). In contrast, earlier studies suggested that NAFLD was more prevalent in women. This higher incidence of NASH may also reflect the higher incidence of metabolic syndrome in morbidly obese men (91.4% vs 76.2%).

Less common in African Americans

In the United States, African Americans have consistently been found to have the lowest prevalence of NAFLD. In a California population study of 159 newly diagnosed NAFLD cases, non-Hispanic whites accounted for 45%, followed by Hispanics (28%), Asians (18%), and African Americans (3%). After controlling for the ethnic composition of the entire cohort, Hispanics had the highest rate of NAFLD and African Americans the lowest.⁶ In Eastern countries such as Japan, the prevalence of NAFLD is estimated to be about 9.3%. Interestingly, about half of the people with NAFLD in Japan were not overweight.⁷

The difference in prevalence of NAFLD in different ethnic groups may be explained by their different rates of metabolic syndrome (21.6% in African Americans vs 23.8% in whites vs 31.9% in Mexican Americans⁸) as well as other genetic and environmental factors.

NAFLD IS USUALLY CLINICALLY SILENT

NAFLD is usually clinically silent, and its impact has most likely been underestimated. Symptoms, if present, are minimal and non-specific, such as fatigue and right upper quadrant discomfort. Most findings on physical examination are also normal. Most patients seek care because of an incidental finding of elevated aminotransferase levels or radiographic studies suggesting the liver is fatty.⁹

The estimated prevalence of aminotransferase elevations in the general population from the third National Health and Nutrition Examination Survey data is 7.9%,¹⁰ with about two-thirds of cases unexplained. Of the unexplained cases, most are strongly associated with metabolic syndrome and probably represent underlying NAFLD.¹⁰

Yet aminotransferase levels are typically normal or elevated by less than five times the upper limit of normal (usually < 250 IU/L).⁹ In contrast to those with alcoholic hepatitis, most patients with NAFLD have a ratio of aspartate aminotransferase to alanine aminotransferase of less than 1. As the disease progresses, the aspartate aminotransferase level increases more than the alanine aminotransferase level, so if the ratio is more than 1, more advanced liver disease may be suspected.¹¹

Levels of other liver enzymes such as alkaline phosphatase and of acute-phase reactants such as ferritin may also be elevated. Ferritin is believed to reflect hepatic injury, inflammation, or insulin resistance.

A DIAGNOSIS OF EXCLUSION

Excessive alcohol consumption must especially be excluded. Most studies defined excessive alcohol consumption as more than 20 to 40 g/day.² Recently, this threshold has been lowered to 20 g/day (roughly two drinks) in men and 10 g/day in women.

Insulin resistance should be estimated, given the close relationship between NAFLD and insulin resistance and the metabolic syndrome. Insulin resistance can be measured accurately in a number of ways. The Homeostasis Model Assessment is an easy method that provides an estimate of insulin resistance based on fasting serum glucose and serum insulin levels.¹³

Serologic tests can rule out hepatitis B and hepatitis C. In those with negative results, especially in those with components of the metabolic syndrome or insulin resistance, NAFLD is responsible for most cases of persistently elevated serum liver enzymes.

Imaging tests

Radiographic evaluation is another noninvasive way to diagnose fatty liver. The sensitivity of either ultrasonography or computed tomography for detecting hepatic steatosis is between 93% and 100% when there is more than 33% fat in the hepatic parenchyma.¹⁴ None of the radiographic methods, including magnetic resonance imaging, can accurately differentiate between nonprogressive simple steatosis and NASH, but the technology is advancing. Contrast ultrasonography and magnetic resonance spectroscopy have shown promise and may become useful in the future.

Other noninvasive tests

Ultrasonographic elastrography (FibroScan), a noninvasive way to measure liver stiffness, has also been used in patients with hepatitis C. Although the preliminary data in NAFLD are interesting, additional validation is needed.

Serum biomarkers, including markers of fibrosis (eg, FibroSURE), apoptosis, and adipocytokines have been used to diagnose NASH. The markers of apoptosis are especially interesting but need further validation.

Liver biopsy remains the gold standard

Because we lack a fully validated noninvasive biomarker of NASH, liver biopsy remains the gold standard for diagnosing it. The minimum histologic criteria for establishing the diagnosis of NASH have been debated; most pathologists require at least 5% hepatic steatosis, mixed lobular inflammation, and hepatocellular ballooning.

In a study of 354 liver biopsies of patients with negative results on serologic tests, NASH was found in 34% and fatty liver in 32%. In the same study, the findings on liver biopsy led to alterations in patient management in 18% of cases.¹⁵

Some clinicians doubt the value of liver biopsy in patients with suspected NASH, in view of possible sampling error in the biopsy specimens (the distribution can be patchy, and if the specimen is taken from an unaffected area, the results can be falsely negative) and because there is no established effective therapy for NAFLD. However, liver biopsy is the only test that can accurately establish the diagnosis of NASH and tell us the stage of liver disease, which has important prognostic implications. Most experts agree that liver biopsy should be considered for patients at risk of advanced liver disease, such as those with persistently elevated liver enzyme levels despite intervention to reverse conditions associated with metabolic syndrome.¹⁶

PATHOGENESIS: THE MULTIPLE-HIT HYPOTHESIS

NAFLD is closely linked to obesity, insulin resistance, and metabolic syndrome.¹³ Insulin allows free fatty esterification and triglyceride fat storage in adipose tissues. When insulin resistance develops, free fatty acids are inappropriately shifted to nonadipose tissues, including the liver. Insulin resistance increases free fatty acid flux to the liver by decreased inhibition of lipolysis and also increased de novo lipogenesis.¹⁷

Insulin resistance and visceral obesity also result in decreased levels of a “protective adipokine,” adiponectin. Adiponectin inhibits liver gluconeogenesis and suppresses lipogenesis. Thus, decreased adiponectin hinders fatty acid oxidation and increases fat accumulation in the liver. Other adipocytokines that are important in NAFLD are resistin, leptin, visfatin, tumor necrosis factor alpha, and interleukin 6.

Apoptosis and oxidative stress may also contribute to the development and progression of NASH. In this context, the “multiple-hit hypothesis” for the pathogenesis of NASH has become quite popular.¹⁸ An in-depth review of the pathogenesis of NAFLD is beyond the scope of this paper; readers are referred to a recently published review on this subject.¹⁹

STEATOSIS IS BENIGN, BUT NASH CAN PROGRESS

Simple steatosis by itself generally has a benign prognosis. In a 1995 cohort study with a median follow-up of 11 years, there was no progression of simple steatosis to NASH or cirrhosis,²⁰ and recent reviews estimate that only a small portion of patients with simple steatosis develop steatohepatitis. The validity of these data is still being debated.

On the other hand, once patients have progressed to NASH, histologic progression has been noted in about 32% to 41% of patients over a median follow-up of 4.3 to 13.7 years.^21,22 This would mean that approximately 9% of patients with NASH may develop cirrhosis.²¹

People with cirrhosis due to NAFLD are at risk of developing liver-related morbidity and of death. In one of the longest follow-up cohort studies (mean follow-up of 13.7 years), end-stage liver disease developed in 5.4%, and hepatocellular carcinoma developed in about 2%. About 20% of the patients died, with more than 70% of the deaths in patients who had NASH at baseline. The survival rate was lower in patients with NASH, whereas no difference in survival was seen in the group with simple steatosis.²²

A number of studies have assessed independent predictors of advanced fibrosis. Most studies suggest that elevated liver enzymes, metabolic syndrome, or type 2 diabetes is associated with advanced liver disease. Although noninvasive biomarkers of fibrosis have been developed for hepatitis C, to date, a fully validated, noninvasive biomarker of fibrosis for NAFLD does not exist.

As noted, the spectrum of NAFLD also includes hepatocellular carcinoma, and in a series of 105 patients with hepatocellular carcinoma, hepatitis C virus accounted for 51% and cryptogenic liver disease accounted for another 29%. Since cases of cryptogenic cirrhosis in the United States are considered to be “burned out NASH,” approximately 13% of patients with hepatocellular carcinoma may have had underlying NAFLD as the cause of their liver disease.²³ These data suggest that, similar to other cirrhotic patients, NAFLD patients with cirrhosis should be screened for hepatocellular carcinoma.

NO CONSENSUS ON TREATMENT

Weight loss

Modest weight loss—less than 2 pounds (1 kg) per week—is associated with a decrease in the incidence of metabolic syndrome and can also improve the histologic features of NASH in more than 80% of cases.²⁴ Loss of as little as 4% to 5% of body weight is also associated with lowering of aminotransferase and fasting insulin levels.²⁵

The mechanism of benefit is via loss of adipose tissue, which decreases insulin resistance. Weight loss by any means, including bariatric surgery for morbid obesity or use of weight-reducing agents, has been correlated with improvement in liver enzyme levels, liver histologic findings, or both.^24,26

However, the traditional low-calorie, low-fat diet may not be optimal for NAFLD patients. In one study,²⁷ patients consuming more than 54% of their calories from carbohydrates compared with those consuming less than 35% had an odds ratio of 6.5 for hepatic inflammation. This finding is not surprising in light of prior research in which high carbohydrate intake increased hepatic de novo lipogenesis. On the other hand, there was no association between total caloric or protein intake and hepatic steatosis or fibrosis. Contrary to traditional beliefs, patients with higher fat intake had less inflammation, steatosis, and fibrosis.

Insulin sensitizers

Given that insulin resistance seems to be the main pathophysiologic culprit in NAFLD, two classes of insulin sensitizers have been studied:

Biguanides act mainly by increasing hepatic insulin sensitivity and reversing insulin resistance induced by tumor necrosis factor alpha.

Glitazones improve insulin sensitivity in both diabetic and euglycemic patients by activating the nuclear transcription factor called peroxisome proliferator-activated receptor (PPAR) gamma.

Both biguanides and glitazones have been found to lower liver enzyme levels, decrease insulin resistance, and improve histopathologic findings. However, the effects of glitazones do not persist after the drugs are stopped, and these drugs and are also associated with an average weight gain of 3 to 6 kg.^28,29

Although these data are encouraging, they are preliminary, and more evidence is needed to establish the safety and efficacy of these drugs in treating patients with NASH.

Antioxidants

Antioxidants such as vitamin E, n-acetyl-l-cysteine, s-adenosylmethionine (SAMe), and betaine have been investigated in the treatment of NAFLD.

Vitamin E has been most widely studied. Being fat-soluble, vitamin E can stabilize mitochondrial function and is theorized to inhibit lipid peroxidation and subsequent free radical reactions. Smaller, nonrandomized trials have found that vitamin E improves biochemical markers of liver inflammation. However, in one of the largest randomized controlled trials (with 45 patients), patients taking vitamin E showed improvement in their fibrosis scores but no differences in their necroinflammatory activity or alanine aminotransferase levels.³⁰ Most studies of antioxidants show at least mild improvement in biochemical or histologic signs of NAFLD.³¹

SAMe and betaine are important antioxidants. However, most studies of SAMe and betaine have been small and inconclusive.

Two large phase III clinical trials are under way at the National Institute of Diabetes and Digestive and Kidney Diseases. They should clarify the role of these agents in the treatment of NASH. The PIVENS (Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients With Nonalcoholic Steatohepatitis) study has completed enrollment of 240 patients, but the final data are not available. The second study, TONIC (Treatment of Nonalcoholic Fatty Liver Disease in Children) will be one of the largest studies of NAFLD in children; it will be looking at vitamin E, metformin, or placebo over a 2-year follow-up. The TONIC study is still under way, so the final data are not yet available.

Ursodeoxycholic acid, another cytoprotective agent, has traditionally been used for primary biliary cirrhosis, but the data are conflicting on its efficacy in NAFLD. Of note, some bile acids are hepatotoxic and facilitate apoptosis via a Fas ligand-mediated pathway. On the other hand, ursodeoxycholic acid is a hydrophilic bile acid that may act to displace the hepatotoxic hydrophobic endogenous bile acids and potentially has an antiapoptotic and cytoprotective effect in NAFLD. Although liver enzyme levels declined in a few of the studies of ursodeoxycholic acid in patients with NAFLD, a large randomized clinical trial (in 166 patients) did not show any significant difference from placebo in liver enzyme levels or liver histologic findings.³²

Lipid-lowering drugs

Lipid-lowering drugs target the high levels of triglycerides and low levels of high-density lipoprotein cholesterol that often occur in insulin resistance and metabolic syndrome associated with NAFLD. A few small studies found that aminotransferase levels fell with both statins and gemfibrozil (Lopid).³³ Even if liver enzyme levels are abnormal, most experts believe that statins are relatively safe to use in patients with NAFLD who need cholesterol-lowering agents. Nevertheless, clinical monitoring of these patients for potential hepatic toxicity is recommended.

Other medications

Other medications, such as pentoxifylline (Pentoxil, Trental), probiotics, and angiotensin-converting enzyme inhibitors, have been used in small studies of patients with NASH, with encouraging but inconclusive results.

Although a number of pilot studies of agents for treating NAFLD have been proposed, they are small and open-label. With the tremendous recent gains in clinical investigation, functional genomics, and proteomics, it is expected that our understanding of NASH and its treatment will be broadened.

In summary, despite the relatively large number of agents tested for the treatment of NAFLD, most of the data are preliminary. Thus, in 2008, there is no established, evidence-based treatment for patients with NASH.

As the nation gets heavier, our livers will get fattier. The prevalence of nonalcoholic fatty liver disease (NAFLD) has been rising in tandem with the rise in obesity ever since the term nonalcoholic steatohepatitis (NASH, a subtype of NAFLD) was coined by Ludwig in 1980.¹ Yet, despite an explosion of research on NAFLD and gains in understanding its epidemiology and pathogenesis, a number of issues remain unresolved, including how to treat it.

NAFLD IS A SPECTRUM

NAFLD is a spectrum. The mildest form is simple fatty liver, or simple steatosis. Next is NASH, or fatty liver with inflammation and evidence of damage to hepatocytes (liver cells). Still more severe is cirrhosis, and in its most extreme form NAFLD can progress to hepatocellular carcinoma or liver failure. The distinction between simple steatosis and NASH is important because their prognoses and management are different.

NAFLD IS COMMON AND LINKED TO OBESITY

NAFLD is the most common cause of elevated liver enzymes and also one of the most common forms of liver disease in the world. It is now estimated to affect about 20% to 30% of people in the United States and other Western countries. In contrast, the prevalence of chronic hepatitis C virus infection is estimated at 3% of the world’s population. In comparison to the prevalence of NAFLD, the prevalence of NASH is much lower: 2% to 3% in the United States.² The incidence of NAFLD is expected to rise further with the increase in obesity in the United States.

NAFLD is even more common in people who are morbidly obese, ie, who have a body mass index greater than 40 kg/m². In a series of studies of morbidly obese patients undergoing bariatric surgery (N = 1,620), the prevalence of hepatic steatosis was 91% (range 85%–98%), and the prevalence of NASH was 37% (range 24%–98%). NASH was not predicted by age or body mass index, but it was more common in men, people with diabetes, and people with insulin resistance.³

Obesity is also increasing in prevalence in children. Since liver biopsies were not done in most pediatric studies, the pediatric prevalence data are based on elevated aminotransferase levels and on ultrasonographic findings of echogenic livers. The overall prevalence of NAFLD in children is estimated at 3% to 10%, but it may be much higher in obese children.⁴

Arun et al⁵ found that the prevalence of NASH in morbidly obese men was almost twice as high as in morbidly obese women (60.3% vs 30.9%). In contrast, earlier studies suggested that NAFLD was more prevalent in women. This higher incidence of NASH may also reflect the higher incidence of metabolic syndrome in morbidly obese men (91.4% vs 76.2%).

Less common in African Americans

In the United States, African Americans have consistently been found to have the lowest prevalence of NAFLD. In a California population study of 159 newly diagnosed NAFLD cases, non-Hispanic whites accounted for 45%, followed by Hispanics (28%), Asians (18%), and African Americans (3%). After controlling for the ethnic composition of the entire cohort, Hispanics had the highest rate of NAFLD and African Americans the lowest.⁶ In Eastern countries such as Japan, the prevalence of NAFLD is estimated to be about 9.3%. Interestingly, about half of the people with NAFLD in Japan were not overweight.⁷

The difference in prevalence of NAFLD in different ethnic groups may be explained by their different rates of metabolic syndrome (21.6% in African Americans vs 23.8% in whites vs 31.9% in Mexican Americans⁸) as well as other genetic and environmental factors.

NAFLD IS USUALLY CLINICALLY SILENT

NAFLD is usually clinically silent, and its impact has most likely been underestimated. Symptoms, if present, are minimal and non-specific, such as fatigue and right upper quadrant discomfort. Most findings on physical examination are also normal. Most patients seek care because of an incidental finding of elevated aminotransferase levels or radiographic studies suggesting the liver is fatty.⁹

The estimated prevalence of aminotransferase elevations in the general population from the third National Health and Nutrition Examination Survey data is 7.9%,¹⁰ with about two-thirds of cases unexplained. Of the unexplained cases, most are strongly associated with metabolic syndrome and probably represent underlying NAFLD.¹⁰

Yet aminotransferase levels are typically normal or elevated by less than five times the upper limit of normal (usually < 250 IU/L).⁹ In contrast to those with alcoholic hepatitis, most patients with NAFLD have a ratio of aspartate aminotransferase to alanine aminotransferase of less than 1. As the disease progresses, the aspartate aminotransferase level increases more than the alanine aminotransferase level, so if the ratio is more than 1, more advanced liver disease may be suspected.¹¹

Levels of other liver enzymes such as alkaline phosphatase and of acute-phase reactants such as ferritin may also be elevated. Ferritin is believed to reflect hepatic injury, inflammation, or insulin resistance.

A DIAGNOSIS OF EXCLUSION

Excessive alcohol consumption must especially be excluded. Most studies defined excessive alcohol consumption as more than 20 to 40 g/day.² Recently, this threshold has been lowered to 20 g/day (roughly two drinks) in men and 10 g/day in women.

Insulin resistance should be estimated, given the close relationship between NAFLD and insulin resistance and the metabolic syndrome. Insulin resistance can be measured accurately in a number of ways. The Homeostasis Model Assessment is an easy method that provides an estimate of insulin resistance based on fasting serum glucose and serum insulin levels.¹³

Serologic tests can rule out hepatitis B and hepatitis C. In those with negative results, especially in those with components of the metabolic syndrome or insulin resistance, NAFLD is responsible for most cases of persistently elevated serum liver enzymes.

Imaging tests

Radiographic evaluation is another noninvasive way to diagnose fatty liver. The sensitivity of either ultrasonography or computed tomography for detecting hepatic steatosis is between 93% and 100% when there is more than 33% fat in the hepatic parenchyma.¹⁴ None of the radiographic methods, including magnetic resonance imaging, can accurately differentiate between nonprogressive simple steatosis and NASH, but the technology is advancing. Contrast ultrasonography and magnetic resonance spectroscopy have shown promise and may become useful in the future.

Other noninvasive tests

Ultrasonographic elastrography (FibroScan), a noninvasive way to measure liver stiffness, has also been used in patients with hepatitis C. Although the preliminary data in NAFLD are interesting, additional validation is needed.

Serum biomarkers, including markers of fibrosis (eg, FibroSURE), apoptosis, and adipocytokines have been used to diagnose NASH. The markers of apoptosis are especially interesting but need further validation.

Liver biopsy remains the gold standard

Because we lack a fully validated noninvasive biomarker of NASH, liver biopsy remains the gold standard for diagnosing it. The minimum histologic criteria for establishing the diagnosis of NASH have been debated; most pathologists require at least 5% hepatic steatosis, mixed lobular inflammation, and hepatocellular ballooning.

In a study of 354 liver biopsies of patients with negative results on serologic tests, NASH was found in 34% and fatty liver in 32%. In the same study, the findings on liver biopsy led to alterations in patient management in 18% of cases.¹⁵

Some clinicians doubt the value of liver biopsy in patients with suspected NASH, in view of possible sampling error in the biopsy specimens (the distribution can be patchy, and if the specimen is taken from an unaffected area, the results can be falsely negative) and because there is no established effective therapy for NAFLD. However, liver biopsy is the only test that can accurately establish the diagnosis of NASH and tell us the stage of liver disease, which has important prognostic implications. Most experts agree that liver biopsy should be considered for patients at risk of advanced liver disease, such as those with persistently elevated liver enzyme levels despite intervention to reverse conditions associated with metabolic syndrome.¹⁶

PATHOGENESIS: THE MULTIPLE-HIT HYPOTHESIS

NAFLD is closely linked to obesity, insulin resistance, and metabolic syndrome.¹³ Insulin allows free fatty esterification and triglyceride fat storage in adipose tissues. When insulin resistance develops, free fatty acids are inappropriately shifted to nonadipose tissues, including the liver. Insulin resistance increases free fatty acid flux to the liver by decreased inhibition of lipolysis and also increased de novo lipogenesis.¹⁷

Insulin resistance and visceral obesity also result in decreased levels of a “protective adipokine,” adiponectin. Adiponectin inhibits liver gluconeogenesis and suppresses lipogenesis. Thus, decreased adiponectin hinders fatty acid oxidation and increases fat accumulation in the liver. Other adipocytokines that are important in NAFLD are resistin, leptin, visfatin, tumor necrosis factor alpha, and interleukin 6.

Apoptosis and oxidative stress may also contribute to the development and progression of NASH. In this context, the “multiple-hit hypothesis” for the pathogenesis of NASH has become quite popular.¹⁸ An in-depth review of the pathogenesis of NAFLD is beyond the scope of this paper; readers are referred to a recently published review on this subject.¹⁹

STEATOSIS IS BENIGN, BUT NASH CAN PROGRESS

Simple steatosis by itself generally has a benign prognosis. In a 1995 cohort study with a median follow-up of 11 years, there was no progression of simple steatosis to NASH or cirrhosis,²⁰ and recent reviews estimate that only a small portion of patients with simple steatosis develop steatohepatitis. The validity of these data is still being debated.

On the other hand, once patients have progressed to NASH, histologic progression has been noted in about 32% to 41% of patients over a median follow-up of 4.3 to 13.7 years.^21,22 This would mean that approximately 9% of patients with NASH may develop cirrhosis.²¹

People with cirrhosis due to NAFLD are at risk of developing liver-related morbidity and of death. In one of the longest follow-up cohort studies (mean follow-up of 13.7 years), end-stage liver disease developed in 5.4%, and hepatocellular carcinoma developed in about 2%. About 20% of the patients died, with more than 70% of the deaths in patients who had NASH at baseline. The survival rate was lower in patients with NASH, whereas no difference in survival was seen in the group with simple steatosis.²²

A number of studies have assessed independent predictors of advanced fibrosis. Most studies suggest that elevated liver enzymes, metabolic syndrome, or type 2 diabetes is associated with advanced liver disease. Although noninvasive biomarkers of fibrosis have been developed for hepatitis C, to date, a fully validated, noninvasive biomarker of fibrosis for NAFLD does not exist.

As noted, the spectrum of NAFLD also includes hepatocellular carcinoma, and in a series of 105 patients with hepatocellular carcinoma, hepatitis C virus accounted for 51% and cryptogenic liver disease accounted for another 29%. Since cases of cryptogenic cirrhosis in the United States are considered to be “burned out NASH,” approximately 13% of patients with hepatocellular carcinoma may have had underlying NAFLD as the cause of their liver disease.²³ These data suggest that, similar to other cirrhotic patients, NAFLD patients with cirrhosis should be screened for hepatocellular carcinoma.

NO CONSENSUS ON TREATMENT

Weight loss

Modest weight loss—less than 2 pounds (1 kg) per week—is associated with a decrease in the incidence of metabolic syndrome and can also improve the histologic features of NASH in more than 80% of cases.²⁴ Loss of as little as 4% to 5% of body weight is also associated with lowering of aminotransferase and fasting insulin levels.²⁵

The mechanism of benefit is via loss of adipose tissue, which decreases insulin resistance. Weight loss by any means, including bariatric surgery for morbid obesity or use of weight-reducing agents, has been correlated with improvement in liver enzyme levels, liver histologic findings, or both.^24,26

However, the traditional low-calorie, low-fat diet may not be optimal for NAFLD patients. In one study,²⁷ patients consuming more than 54% of their calories from carbohydrates compared with those consuming less than 35% had an odds ratio of 6.5 for hepatic inflammation. This finding is not surprising in light of prior research in which high carbohydrate intake increased hepatic de novo lipogenesis. On the other hand, there was no association between total caloric or protein intake and hepatic steatosis or fibrosis. Contrary to traditional beliefs, patients with higher fat intake had less inflammation, steatosis, and fibrosis.

Insulin sensitizers

Given that insulin resistance seems to be the main pathophysiologic culprit in NAFLD, two classes of insulin sensitizers have been studied:

Biguanides act mainly by increasing hepatic insulin sensitivity and reversing insulin resistance induced by tumor necrosis factor alpha.

Glitazones improve insulin sensitivity in both diabetic and euglycemic patients by activating the nuclear transcription factor called peroxisome proliferator-activated receptor (PPAR) gamma.

Both biguanides and glitazones have been found to lower liver enzyme levels, decrease insulin resistance, and improve histopathologic findings. However, the effects of glitazones do not persist after the drugs are stopped, and these drugs and are also associated with an average weight gain of 3 to 6 kg.^28,29

Although these data are encouraging, they are preliminary, and more evidence is needed to establish the safety and efficacy of these drugs in treating patients with NASH.

Antioxidants

Antioxidants such as vitamin E, n-acetyl-l-cysteine, s-adenosylmethionine (SAMe), and betaine have been investigated in the treatment of NAFLD.

Vitamin E has been most widely studied. Being fat-soluble, vitamin E can stabilize mitochondrial function and is theorized to inhibit lipid peroxidation and subsequent free radical reactions. Smaller, nonrandomized trials have found that vitamin E improves biochemical markers of liver inflammation. However, in one of the largest randomized controlled trials (with 45 patients), patients taking vitamin E showed improvement in their fibrosis scores but no differences in their necroinflammatory activity or alanine aminotransferase levels.³⁰ Most studies of antioxidants show at least mild improvement in biochemical or histologic signs of NAFLD.³¹

SAMe and betaine are important antioxidants. However, most studies of SAMe and betaine have been small and inconclusive.

Two large phase III clinical trials are under way at the National Institute of Diabetes and Digestive and Kidney Diseases. They should clarify the role of these agents in the treatment of NASH. The PIVENS (Pioglitazone vs Vitamin E vs Placebo for the Treatment of Non-Diabetic Patients With Nonalcoholic Steatohepatitis) study has completed enrollment of 240 patients, but the final data are not available. The second study, TONIC (Treatment of Nonalcoholic Fatty Liver Disease in Children) will be one of the largest studies of NAFLD in children; it will be looking at vitamin E, metformin, or placebo over a 2-year follow-up. The TONIC study is still under way, so the final data are not yet available.

Ursodeoxycholic acid, another cytoprotective agent, has traditionally been used for primary biliary cirrhosis, but the data are conflicting on its efficacy in NAFLD. Of note, some bile acids are hepatotoxic and facilitate apoptosis via a Fas ligand-mediated pathway. On the other hand, ursodeoxycholic acid is a hydrophilic bile acid that may act to displace the hepatotoxic hydrophobic endogenous bile acids and potentially has an antiapoptotic and cytoprotective effect in NAFLD. Although liver enzyme levels declined in a few of the studies of ursodeoxycholic acid in patients with NAFLD, a large randomized clinical trial (in 166 patients) did not show any significant difference from placebo in liver enzyme levels or liver histologic findings.³²

Lipid-lowering drugs

Lipid-lowering drugs target the high levels of triglycerides and low levels of high-density lipoprotein cholesterol that often occur in insulin resistance and metabolic syndrome associated with NAFLD. A few small studies found that aminotransferase levels fell with both statins and gemfibrozil (Lopid).³³ Even if liver enzyme levels are abnormal, most experts believe that statins are relatively safe to use in patients with NAFLD who need cholesterol-lowering agents. Nevertheless, clinical monitoring of these patients for potential hepatic toxicity is recommended.

Other medications

Other medications, such as pentoxifylline (Pentoxil, Trental), probiotics, and angiotensin-converting enzyme inhibitors, have been used in small studies of patients with NASH, with encouraging but inconclusive results.

Although a number of pilot studies of agents for treating NAFLD have been proposed, they are small and open-label. With the tremendous recent gains in clinical investigation, functional genomics, and proteomics, it is expected that our understanding of NASH and its treatment will be broadened.

In summary, despite the relatively large number of agents tested for the treatment of NAFLD, most of the data are preliminary. Thus, in 2008, there is no established, evidence-based treatment for patients with NASH.

Whether carotid stenting has any advantage over carotid surgery (endarterectomy)—and for which patients—is still a topic of study and debate.

Treatment of carotid atherosclerosis and stenosis is important in preventing stroke and its comorbidities. Today, three main treatments exist: medical management (lipid-lowering, antihypertensive, and antiplatelet therapy), surgery, and, more recently, carotid angioplasty and stenting. The rationale for these treatments is to decrease the risk of cerebral infarction by stabilizing or removing plaque and improving blood flow.

Surgery has proven beneficial in patients with symptomatic carotid stenosis greater than 50% or asymptomatic stenosis greater than 60%, but it is risky in some patients. Stenting has evolved in part from the success of surgery and the need for alternative treatments for patients who are at unacceptable risk of perioperative complications. However, it does not have a clear advantage over surgery in patients at average risk. Further, its use in patients with asymptomatic stenosis of any severity is still controversial.

In this paper we review the major trials of carotid endarterectomy and stenting and summarize what we know today about who should undergo these therapies.

NOT ALL STROKES ARE DUE TO CAROTID ATHEROSCLEROSIS

Depending on the institution’s referral pattern and population served, between 80% and 90% of strokes are ischemic (the rest being hemorrhagic).¹ Atherosclerosis of large arteries (typically defined as more than 50% stenosis of a major brain artery or branch cortical artery²) is just one cause of ischemic stroke, but it is an important one. Other identifiable causes of ischemic stroke include cardioembolism and small-artery occlusion (lacunar stroke), and some cases are idiopathic.

Large-artery atherosclerotic disease can damage the brain gradually, with carotid stenosis resulting in hypoperfusion and subsequent cerebral infarction. More commonly, however, the carotid plaque often seen in large-artery atherosclerotic disease can ulcerate and occlude the vessel acutely or generate platelet aggregates that may embolize, resulting in cerebral infarction or transient ischemic attack.

In the Lausanne Stroke Registry,³ the rate of ischemic stroke in patients with a greater than 50% large-artery stenosis ranged from 27% in 1979 to 17% in 2003, the decline likely being due to therapeutic advances.

SURGERY BEATS MEDICAL THERAPY FOR CAROTID ATHEROSCLEROSIS

Four landmark trials provided substantial evidence that carotid endarterectomy is better than medical management in patients with symptomatic or asymptomatic high-grade stenosis. These trials indirectly paved the way for carotid stenting.

The North American Symptomatic Carotid Endarterectomy Trial (NASCET)

Patients at 50 clinical centers who had had a hemispheric or retinal transient ischemic attack or a nondisabling stroke were randomized to undergo surgery (carotid endarterectomy) or no surgery. All patients received maximal medical management consisting of blood pressure control, lipid management if indicated, and antiplatelet therapy with aspirin. At baseline, 37% of patients were taking 650 mg or more of aspirin per day, and 11% were taking less than 325 mg per day. The patients were stratified into two prespecified groups on the basis of the severity of carotid stenosis: those with narrowing of 30% to 69% and those with narrowing of 70% to 99%.

Results in high-grade stenosis. In August 1991, the investigators published their results in patients with symptomatic high-grade (70%–99%) stenosis.⁴ Surgical treatment was more beneficial than medical management alone: the cumulative risk of any ipsilateral stroke at 2 years was 26% in the medical group and 9% in the surgical group, an absolute risk reduction of 17%. The benefit of endarterectomy was still apparent at 8 years of follow-up.⁵

Results in moderate stenosis. In 1998, the investigators published their results in patients with symptomatic moderate (< 70%) stenosis.⁵ Surgery was more beneficial than medical therapy in this subgroup as well: at 5 years, the rate of any ipsilateral stroke in patients with 50% to 69% stenosis was 15.7% in those treated surgically and 22.2% in those treated medically (P = .045). In patients with less than 50% stenosis, the 5-year stroke rate was not significantly lower with endarterectomy than with medical therapy.

The European Carotid Surgery Trial (ECST)

The ECST,⁶ published in 1998, corroborated the NASCET findings. This multicenter, randomized, controlled trial enrolled 3,024 patients with symptoms of at least one transient ischemic attack in the distribution of one or both carotid arteries.

Results. In patients with stenosis of greater than 80% (60% by the NASCET criteria for calculating angiographic stenosis), the frequency of major stroke or death at 3 years was 26.5% in the control group and 14.9% in the surgery group, an absolute difference of 11.6%.

The Endarterectomy for Asymptomatic Carotid Artery Stenosis (ACAS) trial

The NASCET and ECST studies made it clear that select groups of patients with symptomatic carotid stenosis benefit from carotid endarterectomy. But what about patients with stenosis but no prior stroke?

The ACAS trial aimed to find out.⁷ In this pivotal study, 1,662 patients with asymptomatic carotid artery stenosis greater than 60% were randomized to receive either medical therapy alone or medical plus surgical therapy.

Results were published in 2004. After a median follow-up of 2.7 years, the aggregate 5-year risk of ipsilateral stroke, any perioperative stroke, or death was estimated to be 5.1% in the surgical group and 11.0% in the medical group, a relative risk reduction of 53%. However, for surgery to be beneficial, the rate of perioperative death and other serious complications had to be less than 3%, and the expected patient survival had to be at least 5 years.

Of note, the benefit of carotid endarterectomy in this study was predominantly in men, with less of a benefit for women and diabetic patients. Furthermore, even though endarterectomy was beneficial in this asymptomatic cohort, the overall benefit in terms of stroke risk reduction was small compared with that in NASCET and ECST, in which patients had symptomatic disease.

The Asymptomatic Carotid Surgery Trial (ACST)

In this European version of ACAS, published in 2004, 3,120 patients with asymptomatic carotid narrowing on ultrasonography were randomized to undergo surgery or medical therapy.

Results. The risk of stroke or death within 30 days of carotid endarterectomy was 3.1%. In patients younger than 75 years who had carotid narrowing of 70% or more, immediate surgery decreased the net 5-year stroke risk from 12% to 6%.⁸

WHO SHOULD NOT UNDERGO CAROTID ENDARTERECTOMY?

From these studies, we can conclude that patients with symptomatic carotid stenosis of 50% or greater and patients with asymptomatic stenosis of 60% or greater benefit from carotid endarterectomy, but only if the perioperative rate of death and other serious complications is less than 3%.⁷

What are the risk factors for complications during this surgery? In 2006, Cremonesi et al,⁹ in a consensus paper, defined patients as being at high risk if they had any of the following:

• Contralateral laryngeal nerve palsy
• Radiation therapy to the neck
• Previous carotid endarterectomy with recurrent stenosis
• Lesions high in the cervical internal carotid artery or below the clavicle in the common carotid artery
• Severe tandem lesions
• Age greater than 80 years
• Severe pulmonary disease
• Congestive heart failure (New York Heart
• Association class 3 or 4) or known severe left ventricular dysfunction
• Open heart surgery needed within 6 weeks
• Myocardial infarction within the past 4 weeks
• Unstable angina
• Contralateral carotid occlusion.

Could endovascular treatment be the answer for these patients at high risk who should not undergo carotid endarterectomy? Indeed, the procedure is being studied extensively and performed more frequently. We summarize the major studies below.

STUDIES OF CAROTID STENTING VS ENDARTERECTOMY

The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS)

This study, published in 2001,¹⁰ was the first randomized, multicenter trial to compare the risks and benefits of endovascular treatment (angioplasty with or without stenting) of carotid and vertebral artery stenosis with those of conventional surgery.

To be included, patients had to have carotid artery stenosis (symptomatic or asymptomatic) that was suitable for either carotid endarterectomy or endovascular treatment. Patients were not grouped on the basis of the severity of their stenosis, but the mean stenosis in randomized patients was 86%.

A total of 504 patients were enrolled, of whom 251 were randomized to undergo endovascular treatment. Most patients in this group underwent angioplasty alone, but 26% also received stents because of suboptimal vessel dilatation or at the discretion of the intervening physician.

The primary end point was any disabling stroke or death. Secondary end points were any ipsilateral stroke lasting longer than 7 days and the combination of death or disabling ipsilateral stroke.

The results showed no significant difference between endovascular treatment and surgery in any of these end points at 3 years. However, the overall rates of procedural stroke and death were nearly double those seen in NASCET and ECST. The investigators could not determine the reason for this higher risk, but they hypothesized that CAVATAS included patients at higher risk.

The restenosis rate was higher in the endovascular therapy group (14%) than in the surgery group (4%; P < .001). On the other hand, the surgery group had a higher rate of minor complications, including cranial nerve palsies and neck hematomas.

Carotid Revascularization With Endarterectomy or Stenting Systems (CARESS)

This prospective, multicenter, phase 2 trial, published in 2003, compared the outcomes of standard carotid endarterectomy vs carotid artery stenting using distal embolic protection devices.¹¹ All the patients in this study had at least 50% symptomatic stenosis or 75% asymptomatic stenosis.

Results. At 30 days, 7 (2.4%) of 254 patients in the endarterectomy group had had strokes, and one of the 7 patients with stroke died, so the combined rate of stroke or death (the primary end point) was 2.4%. In the stenting group, 3 (2.1%) of 143 patients had strokes and no patients died. Overall, there was no significant difference in the composite of death, stroke, or myocardial infarction (the secondary end point): 3% for carotid endarterectomy and 2% for stenting patients.

The Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial

In this trial,¹² published in 2004, patients had to have either symptomatic carotid disease with 50% stenosis or greater or asymptomatic stenosis of 80% or greater, determined by ultrasonography. Further, all patients had to have at least one comorbid condition that increased their perioperative risk. Up until this point, no trial had strictly defined patients at increased risk for complications after carotid endarterectomy and assessed subsequent outcomes. The risk factors included severe cardiac or pulmonary disease, age greater than 80, postendarterectomy carotid stenosis, previous neck surgery, previous neck radiation, contralateral recurrent laryngeal nerve palsy, and contralateral carotid occlusion.

Patients were randomized to undergo carotid artery stenting with distal protection or carotid endarterectomy.

The primary end points of this study were the cumulative incidence of major cardiovascular events at 1 year; death, stroke, or myocardial infarction within 30 days of intervention; and ipsilateral stroke between 31 days and 1 year. Secondary outcomes measured were the rates of target-vessel recanalization at 1 year, cranial nerve palsy, and surgical site complications.

Results. The rate of stroke or death was similar in both groups. The stenting group had fewer adverse cardiac events (mainly non-Q-wave myocardial infarction) than the surgery group. At 1 year the rate of major ipsilateral stroke was 3.3% in the endarterectomy group vs 0% in the stenting group (the difference was not significant), and the cardiovascular event rates continued to be higher in the endarterectomy group.

The investigators noted that myocardial infarction was included as a primary end point because patients with atherosclerotic vascular disease who undergo either stenting or endarterectomy are at a substantial risk of myocardial infarction, and a Q-wave or a non-Q-wave myocardial infarction in the perioperative period increases the risk of future complications and death. A perioperative non-Q-wave infarction increases the risk of death by a factor of 6 and increases the risk of myocardial infarction by a factor of 27 in the subsequent 6 months.

Overall, this study presents evidence that stenting, using distal embolic protection devices, is not inferior to endarterectomy and has fewer cardiovascular complications in patients who have at least one risk factor.

This recent multicenter, randomized study¹³ was designed to determine if stenting is as good as (not inferior to) carotid endarterectomy in patients with symptomatic carotid stenosis of at least 60%. The primary end point was to be the incidence of stroke or death within 30 days after treatment. However, the trial was stopped early after the inclusion of 527 patients for reasons of safety and futility.

Results. The 30-day incidence of any stroke or death was higher in the stenting group (9.6% vs 3.9%). The relative risk of any stroke or death after stenting as compared with endarterectomy was 2.5. The 30-day incidence of disabling stroke or death was also higher in the stenting group (3.4% vs 1.5%; relative risk 2.2). At 6 months, the incidence of any stroke or death was 6.1% after endarterectomy and 11.7% after stenting (P = .02). There was a trend toward more major local complications after stenting and systemic complications after endarterectomy. Cranial-nerve injury was more common after endarterectomy than after stenting (as expected). Overall, death and stroke rates were lower at 1 month and 6 months with endarterectomy than with stenting.

The Stent-Protected Angioplasty Versus Carotid Endarterectomy (SPACE) trial

This randomized, multicenter study,¹⁴ published in 2006, was also designed to compare the safety and efficacy of carotid stenting and endarterectomy. Some 1,200 patients with symptomatic carotid artery stenosis confirmed by ultrasonography were randomly assigned within 180 days of a transient ischemic attack or moderate stroke to undergo carotid artery stenting (n = 605) or carotid endarterectomy (n = 595). The primary end point was ipsilateral ischemic stroke or death 30 days after the procedure. A total of 1,183 patients were included in the analysis.

Results. The rate of the primary end point was 6.84% with stenting and 6.34% with endarterectomy. The study failed to prove the noninferiority of carotid artery stenting compared with carotid endarterectomy for the periprocedural complication rate. Results at 6 to 24 months are awaited.

The Carotid Revascularization Endarterectomy Versus Stenting (CREST) trial

Perhaps the most anxiously awaited results are those of the CREST trial,¹⁵ funded by the National Institutes of Health. This is a prospective, randomized, parallel, two-arm, multicenter clinical trial with blinded end point evaluation. Anticipated enrollment will include 2,500 patients. Patients are eligible for enrollment if they have symptoms of carotid stenosis within 180 days of a stroke or transient ischemic attack with ipsilateral carotid stenosis of at least 50% by angiography (70% by ultrasonography), or if they have asymptomatic carotid stenosis of at least 60% by angiography (70% by ultrasonography).

Patients are being randomized to undergo either carotid artery stenting or carotid endarterectomy. All receive aspirin as anti-platelet therapy, treatment for hypertension, and management of other stroke risk factors. Follow-up will last 4 years, with clinic visits at 1, 6, 12, 18, 24, 30, 36, 42, and 48 months. Primary outcome measures will be rates of death, stroke, or myocardial infarction at 30 days postoperatively, and ipsilateral stroke at 30 days postoperatively.

As of February 2007, 1,506 patients had been enrolled and 1,453 had been randomized at 94 sites in North America.

MEDICAID AND MEDICARE NOW PAY FOR THESE THERAPIES

An important practical consideration for patients and physicians is whether Medicaid and Medicare will pay for these therapies.

In July 2001, Medicare began to cover percutaneous transluminal angioplasty of the carotid artery with concurrent stent placement, when furnished in accordance with US Food and Drug Administration (FDA) protocols governing Category B (nonexperimental) investigational device exemption clinical trials.¹⁶ Angioplasty of the carotid artery, when provided solely for the purpose of carotid artery dilation concurrent with carotid stent placement, is considered to be a reasonable and necessary service when provided in the context of clinical trials.

In March 2005, Medicare began to provide coverage for percutaneous transluminal angioplasty of the carotid artery concurrent with the placement of an FDA-approved carotid stent with embolic protection for the following groups of patients:

• Those who would be at high risk during carotid endarterectomy and who also have symptomatic carotid artery stenosis of 70% or greater. Coverage is limited to procedures performed using FDA-approved carotid artery stenting systems and embolic protection devices.
• Those who would be at high risk during endarterectomy and who have symptomatic carotid artery stenosis of 50% to 70%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting post-approval.
• Those who would be at high risk during carotid endarterectomy and have asymptomatic carotid artery stenosis greater than 80%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting postapproval studies.

As noted above, Medicare and Medicaid will only cover carotid stenting if the stent system is FDA-approved, with concomitant use of a distal embolic protection device. However, in view of conflicting data from stenting trials to date, including EVA-3S¹³ and SPACE,¹⁴ it remains to be seen if emboli protection devices significantly reduce periprocedural stroke rates. The FDA recommends that if it is not technically possible to use one of these devices, then the procedure should be aborted due to safety issues.

These coverage decisions are an important practical aspect of carotid stenting and they should be familiar to physicians when they see and refer patients with carotid disease.

WHAT CAN WE SAY AT THIS POINT?

Given the multiple recent and ongoing trials of stenting vs endarterectomy in carotid stenosis, debate continues as to what the role of stenting will be in the future. What can we say at this point?

In patients with asymptomatic carotid stenosis of greater than 60% or symptomatic carotid stenosis of greater than 50%, carotid endarterectomy has been proven to be superior to medical therapy alone.

The efficacy and safety of carotid stenting compared with carotid endarterectomy is still uncertain. In the trials reviewed above, carotid stenting did not appear to have a clear advantage over endarterectomy in patients at average surgical risk. Stenting may be most advantageous when used in patients with symptomatic carotid stenosis who would be at high operative risk, as indicated by the SAPPHIRE trial.

In patients with severe but asymptomatic carotid stenosis who are at high operative risk, the addition of carotid angioplasty and stenting to maximum medical therapy remains controversial. The periprocedural complication rate in these patients may actually exceed the rate of stroke in asymptomatic patients with greater than 60% stenosis who do not undergo stenting or surgery. In addition, subgroup analyses of patients with 70% to 99% symptomatic stenosis in various trials show that surgical benefit is greater in men than in women, and it remains to be seen whether there is any benefit in women with moderate stenoses, asymptomatic lesions, or both.¹⁷

Further experience and study are needed, and the results of the Carotid Stenting vs Surgery of Severe Carotid Artery Disease and Stroke Prevention in Asymptomatic Patients (ACT I) study (comparing stenting and surgery in asymptomatic carotid stenosis), and the ongoing CREST trial (comparing stenting and surgery in symptomatic and asymptomatic carotid stenosis) are eagerly awaited. Until then, clinicians should continue to weigh individual patient risks and benefits when referring patients for surgical treatment of carotid athero-sclerotic disease. Regardless of whether surgery is undertaken, maximal medical therapy with the use of antiplatelet agents, blood pressure control, and statin therapy remains the mainstay of treatment.

Whether carotid stenting has any advantage over carotid surgery (endarterectomy)—and for which patients—is still a topic of study and debate.

Treatment of carotid atherosclerosis and stenosis is important in preventing stroke and its comorbidities. Today, three main treatments exist: medical management (lipid-lowering, antihypertensive, and antiplatelet therapy), surgery, and, more recently, carotid angioplasty and stenting. The rationale for these treatments is to decrease the risk of cerebral infarction by stabilizing or removing plaque and improving blood flow.

Surgery has proven beneficial in patients with symptomatic carotid stenosis greater than 50% or asymptomatic stenosis greater than 60%, but it is risky in some patients. Stenting has evolved in part from the success of surgery and the need for alternative treatments for patients who are at unacceptable risk of perioperative complications. However, it does not have a clear advantage over surgery in patients at average risk. Further, its use in patients with asymptomatic stenosis of any severity is still controversial.

In this paper we review the major trials of carotid endarterectomy and stenting and summarize what we know today about who should undergo these therapies.

NOT ALL STROKES ARE DUE TO CAROTID ATHEROSCLEROSIS

Depending on the institution’s referral pattern and population served, between 80% and 90% of strokes are ischemic (the rest being hemorrhagic).¹ Atherosclerosis of large arteries (typically defined as more than 50% stenosis of a major brain artery or branch cortical artery²) is just one cause of ischemic stroke, but it is an important one. Other identifiable causes of ischemic stroke include cardioembolism and small-artery occlusion (lacunar stroke), and some cases are idiopathic.

Large-artery atherosclerotic disease can damage the brain gradually, with carotid stenosis resulting in hypoperfusion and subsequent cerebral infarction. More commonly, however, the carotid plaque often seen in large-artery atherosclerotic disease can ulcerate and occlude the vessel acutely or generate platelet aggregates that may embolize, resulting in cerebral infarction or transient ischemic attack.

In the Lausanne Stroke Registry,³ the rate of ischemic stroke in patients with a greater than 50% large-artery stenosis ranged from 27% in 1979 to 17% in 2003, the decline likely being due to therapeutic advances.

SURGERY BEATS MEDICAL THERAPY FOR CAROTID ATHEROSCLEROSIS

Four landmark trials provided substantial evidence that carotid endarterectomy is better than medical management in patients with symptomatic or asymptomatic high-grade stenosis. These trials indirectly paved the way for carotid stenting.

The North American Symptomatic Carotid Endarterectomy Trial (NASCET)

Patients at 50 clinical centers who had had a hemispheric or retinal transient ischemic attack or a nondisabling stroke were randomized to undergo surgery (carotid endarterectomy) or no surgery. All patients received maximal medical management consisting of blood pressure control, lipid management if indicated, and antiplatelet therapy with aspirin. At baseline, 37% of patients were taking 650 mg or more of aspirin per day, and 11% were taking less than 325 mg per day. The patients were stratified into two prespecified groups on the basis of the severity of carotid stenosis: those with narrowing of 30% to 69% and those with narrowing of 70% to 99%.

Results in high-grade stenosis. In August 1991, the investigators published their results in patients with symptomatic high-grade (70%–99%) stenosis.⁴ Surgical treatment was more beneficial than medical management alone: the cumulative risk of any ipsilateral stroke at 2 years was 26% in the medical group and 9% in the surgical group, an absolute risk reduction of 17%. The benefit of endarterectomy was still apparent at 8 years of follow-up.⁵

Results in moderate stenosis. In 1998, the investigators published their results in patients with symptomatic moderate (< 70%) stenosis.⁵ Surgery was more beneficial than medical therapy in this subgroup as well: at 5 years, the rate of any ipsilateral stroke in patients with 50% to 69% stenosis was 15.7% in those treated surgically and 22.2% in those treated medically (P = .045). In patients with less than 50% stenosis, the 5-year stroke rate was not significantly lower with endarterectomy than with medical therapy.

The European Carotid Surgery Trial (ECST)

The ECST,⁶ published in 1998, corroborated the NASCET findings. This multicenter, randomized, controlled trial enrolled 3,024 patients with symptoms of at least one transient ischemic attack in the distribution of one or both carotid arteries.

Results. In patients with stenosis of greater than 80% (60% by the NASCET criteria for calculating angiographic stenosis), the frequency of major stroke or death at 3 years was 26.5% in the control group and 14.9% in the surgery group, an absolute difference of 11.6%.

The Endarterectomy for Asymptomatic Carotid Artery Stenosis (ACAS) trial

The NASCET and ECST studies made it clear that select groups of patients with symptomatic carotid stenosis benefit from carotid endarterectomy. But what about patients with stenosis but no prior stroke?

The ACAS trial aimed to find out.⁷ In this pivotal study, 1,662 patients with asymptomatic carotid artery stenosis greater than 60% were randomized to receive either medical therapy alone or medical plus surgical therapy.

Results were published in 2004. After a median follow-up of 2.7 years, the aggregate 5-year risk of ipsilateral stroke, any perioperative stroke, or death was estimated to be 5.1% in the surgical group and 11.0% in the medical group, a relative risk reduction of 53%. However, for surgery to be beneficial, the rate of perioperative death and other serious complications had to be less than 3%, and the expected patient survival had to be at least 5 years.

Of note, the benefit of carotid endarterectomy in this study was predominantly in men, with less of a benefit for women and diabetic patients. Furthermore, even though endarterectomy was beneficial in this asymptomatic cohort, the overall benefit in terms of stroke risk reduction was small compared with that in NASCET and ECST, in which patients had symptomatic disease.

The Asymptomatic Carotid Surgery Trial (ACST)

In this European version of ACAS, published in 2004, 3,120 patients with asymptomatic carotid narrowing on ultrasonography were randomized to undergo surgery or medical therapy.

Results. The risk of stroke or death within 30 days of carotid endarterectomy was 3.1%. In patients younger than 75 years who had carotid narrowing of 70% or more, immediate surgery decreased the net 5-year stroke risk from 12% to 6%.⁸

WHO SHOULD NOT UNDERGO CAROTID ENDARTERECTOMY?

From these studies, we can conclude that patients with symptomatic carotid stenosis of 50% or greater and patients with asymptomatic stenosis of 60% or greater benefit from carotid endarterectomy, but only if the perioperative rate of death and other serious complications is less than 3%.⁷

What are the risk factors for complications during this surgery? In 2006, Cremonesi et al,⁹ in a consensus paper, defined patients as being at high risk if they had any of the following:

• Contralateral laryngeal nerve palsy
• Radiation therapy to the neck
• Previous carotid endarterectomy with recurrent stenosis
• Lesions high in the cervical internal carotid artery or below the clavicle in the common carotid artery
• Severe tandem lesions
• Age greater than 80 years
• Severe pulmonary disease
• Congestive heart failure (New York Heart
• Association class 3 or 4) or known severe left ventricular dysfunction
• Open heart surgery needed within 6 weeks
• Myocardial infarction within the past 4 weeks
• Unstable angina
• Contralateral carotid occlusion.

Could endovascular treatment be the answer for these patients at high risk who should not undergo carotid endarterectomy? Indeed, the procedure is being studied extensively and performed more frequently. We summarize the major studies below.

STUDIES OF CAROTID STENTING VS ENDARTERECTOMY

The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS)

This study, published in 2001,¹⁰ was the first randomized, multicenter trial to compare the risks and benefits of endovascular treatment (angioplasty with or without stenting) of carotid and vertebral artery stenosis with those of conventional surgery.

To be included, patients had to have carotid artery stenosis (symptomatic or asymptomatic) that was suitable for either carotid endarterectomy or endovascular treatment. Patients were not grouped on the basis of the severity of their stenosis, but the mean stenosis in randomized patients was 86%.

A total of 504 patients were enrolled, of whom 251 were randomized to undergo endovascular treatment. Most patients in this group underwent angioplasty alone, but 26% also received stents because of suboptimal vessel dilatation or at the discretion of the intervening physician.

The primary end point was any disabling stroke or death. Secondary end points were any ipsilateral stroke lasting longer than 7 days and the combination of death or disabling ipsilateral stroke.

The results showed no significant difference between endovascular treatment and surgery in any of these end points at 3 years. However, the overall rates of procedural stroke and death were nearly double those seen in NASCET and ECST. The investigators could not determine the reason for this higher risk, but they hypothesized that CAVATAS included patients at higher risk.

The restenosis rate was higher in the endovascular therapy group (14%) than in the surgery group (4%; P < .001). On the other hand, the surgery group had a higher rate of minor complications, including cranial nerve palsies and neck hematomas.

Carotid Revascularization With Endarterectomy or Stenting Systems (CARESS)

This prospective, multicenter, phase 2 trial, published in 2003, compared the outcomes of standard carotid endarterectomy vs carotid artery stenting using distal embolic protection devices.¹¹ All the patients in this study had at least 50% symptomatic stenosis or 75% asymptomatic stenosis.

Results. At 30 days, 7 (2.4%) of 254 patients in the endarterectomy group had had strokes, and one of the 7 patients with stroke died, so the combined rate of stroke or death (the primary end point) was 2.4%. In the stenting group, 3 (2.1%) of 143 patients had strokes and no patients died. Overall, there was no significant difference in the composite of death, stroke, or myocardial infarction (the secondary end point): 3% for carotid endarterectomy and 2% for stenting patients.

The Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial

In this trial,¹² published in 2004, patients had to have either symptomatic carotid disease with 50% stenosis or greater or asymptomatic stenosis of 80% or greater, determined by ultrasonography. Further, all patients had to have at least one comorbid condition that increased their perioperative risk. Up until this point, no trial had strictly defined patients at increased risk for complications after carotid endarterectomy and assessed subsequent outcomes. The risk factors included severe cardiac or pulmonary disease, age greater than 80, postendarterectomy carotid stenosis, previous neck surgery, previous neck radiation, contralateral recurrent laryngeal nerve palsy, and contralateral carotid occlusion.

Patients were randomized to undergo carotid artery stenting with distal protection or carotid endarterectomy.

The primary end points of this study were the cumulative incidence of major cardiovascular events at 1 year; death, stroke, or myocardial infarction within 30 days of intervention; and ipsilateral stroke between 31 days and 1 year. Secondary outcomes measured were the rates of target-vessel recanalization at 1 year, cranial nerve palsy, and surgical site complications.

Results. The rate of stroke or death was similar in both groups. The stenting group had fewer adverse cardiac events (mainly non-Q-wave myocardial infarction) than the surgery group. At 1 year the rate of major ipsilateral stroke was 3.3% in the endarterectomy group vs 0% in the stenting group (the difference was not significant), and the cardiovascular event rates continued to be higher in the endarterectomy group.

The investigators noted that myocardial infarction was included as a primary end point because patients with atherosclerotic vascular disease who undergo either stenting or endarterectomy are at a substantial risk of myocardial infarction, and a Q-wave or a non-Q-wave myocardial infarction in the perioperative period increases the risk of future complications and death. A perioperative non-Q-wave infarction increases the risk of death by a factor of 6 and increases the risk of myocardial infarction by a factor of 27 in the subsequent 6 months.

Overall, this study presents evidence that stenting, using distal embolic protection devices, is not inferior to endarterectomy and has fewer cardiovascular complications in patients who have at least one risk factor.

This recent multicenter, randomized study¹³ was designed to determine if stenting is as good as (not inferior to) carotid endarterectomy in patients with symptomatic carotid stenosis of at least 60%. The primary end point was to be the incidence of stroke or death within 30 days after treatment. However, the trial was stopped early after the inclusion of 527 patients for reasons of safety and futility.

Results. The 30-day incidence of any stroke or death was higher in the stenting group (9.6% vs 3.9%). The relative risk of any stroke or death after stenting as compared with endarterectomy was 2.5. The 30-day incidence of disabling stroke or death was also higher in the stenting group (3.4% vs 1.5%; relative risk 2.2). At 6 months, the incidence of any stroke or death was 6.1% after endarterectomy and 11.7% after stenting (P = .02). There was a trend toward more major local complications after stenting and systemic complications after endarterectomy. Cranial-nerve injury was more common after endarterectomy than after stenting (as expected). Overall, death and stroke rates were lower at 1 month and 6 months with endarterectomy than with stenting.

The Stent-Protected Angioplasty Versus Carotid Endarterectomy (SPACE) trial

This randomized, multicenter study,¹⁴ published in 2006, was also designed to compare the safety and efficacy of carotid stenting and endarterectomy. Some 1,200 patients with symptomatic carotid artery stenosis confirmed by ultrasonography were randomly assigned within 180 days of a transient ischemic attack or moderate stroke to undergo carotid artery stenting (n = 605) or carotid endarterectomy (n = 595). The primary end point was ipsilateral ischemic stroke or death 30 days after the procedure. A total of 1,183 patients were included in the analysis.

Results. The rate of the primary end point was 6.84% with stenting and 6.34% with endarterectomy. The study failed to prove the noninferiority of carotid artery stenting compared with carotid endarterectomy for the periprocedural complication rate. Results at 6 to 24 months are awaited.

The Carotid Revascularization Endarterectomy Versus Stenting (CREST) trial

Perhaps the most anxiously awaited results are those of the CREST trial,¹⁵ funded by the National Institutes of Health. This is a prospective, randomized, parallel, two-arm, multicenter clinical trial with blinded end point evaluation. Anticipated enrollment will include 2,500 patients. Patients are eligible for enrollment if they have symptoms of carotid stenosis within 180 days of a stroke or transient ischemic attack with ipsilateral carotid stenosis of at least 50% by angiography (70% by ultrasonography), or if they have asymptomatic carotid stenosis of at least 60% by angiography (70% by ultrasonography).

Patients are being randomized to undergo either carotid artery stenting or carotid endarterectomy. All receive aspirin as anti-platelet therapy, treatment for hypertension, and management of other stroke risk factors. Follow-up will last 4 years, with clinic visits at 1, 6, 12, 18, 24, 30, 36, 42, and 48 months. Primary outcome measures will be rates of death, stroke, or myocardial infarction at 30 days postoperatively, and ipsilateral stroke at 30 days postoperatively.

As of February 2007, 1,506 patients had been enrolled and 1,453 had been randomized at 94 sites in North America.

MEDICAID AND MEDICARE NOW PAY FOR THESE THERAPIES

An important practical consideration for patients and physicians is whether Medicaid and Medicare will pay for these therapies.

In July 2001, Medicare began to cover percutaneous transluminal angioplasty of the carotid artery with concurrent stent placement, when furnished in accordance with US Food and Drug Administration (FDA) protocols governing Category B (nonexperimental) investigational device exemption clinical trials.¹⁶ Angioplasty of the carotid artery, when provided solely for the purpose of carotid artery dilation concurrent with carotid stent placement, is considered to be a reasonable and necessary service when provided in the context of clinical trials.

In March 2005, Medicare began to provide coverage for percutaneous transluminal angioplasty of the carotid artery concurrent with the placement of an FDA-approved carotid stent with embolic protection for the following groups of patients:

• Those who would be at high risk during carotid endarterectomy and who also have symptomatic carotid artery stenosis of 70% or greater. Coverage is limited to procedures performed using FDA-approved carotid artery stenting systems and embolic protection devices.
• Those who would be at high risk during endarterectomy and who have symptomatic carotid artery stenosis of 50% to 70%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting post-approval.
• Those who would be at high risk during carotid endarterectomy and have asymptomatic carotid artery stenosis greater than 80%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting postapproval studies.

As noted above, Medicare and Medicaid will only cover carotid stenting if the stent system is FDA-approved, with concomitant use of a distal embolic protection device. However, in view of conflicting data from stenting trials to date, including EVA-3S¹³ and SPACE,¹⁴ it remains to be seen if emboli protection devices significantly reduce periprocedural stroke rates. The FDA recommends that if it is not technically possible to use one of these devices, then the procedure should be aborted due to safety issues.

These coverage decisions are an important practical aspect of carotid stenting and they should be familiar to physicians when they see and refer patients with carotid disease.

WHAT CAN WE SAY AT THIS POINT?

Given the multiple recent and ongoing trials of stenting vs endarterectomy in carotid stenosis, debate continues as to what the role of stenting will be in the future. What can we say at this point?

In patients with asymptomatic carotid stenosis of greater than 60% or symptomatic carotid stenosis of greater than 50%, carotid endarterectomy has been proven to be superior to medical therapy alone.

The efficacy and safety of carotid stenting compared with carotid endarterectomy is still uncertain. In the trials reviewed above, carotid stenting did not appear to have a clear advantage over endarterectomy in patients at average surgical risk. Stenting may be most advantageous when used in patients with symptomatic carotid stenosis who would be at high operative risk, as indicated by the SAPPHIRE trial.

In patients with severe but asymptomatic carotid stenosis who are at high operative risk, the addition of carotid angioplasty and stenting to maximum medical therapy remains controversial. The periprocedural complication rate in these patients may actually exceed the rate of stroke in asymptomatic patients with greater than 60% stenosis who do not undergo stenting or surgery. In addition, subgroup analyses of patients with 70% to 99% symptomatic stenosis in various trials show that surgical benefit is greater in men than in women, and it remains to be seen whether there is any benefit in women with moderate stenoses, asymptomatic lesions, or both.¹⁷

Further experience and study are needed, and the results of the Carotid Stenting vs Surgery of Severe Carotid Artery Disease and Stroke Prevention in Asymptomatic Patients (ACT I) study (comparing stenting and surgery in asymptomatic carotid stenosis), and the ongoing CREST trial (comparing stenting and surgery in symptomatic and asymptomatic carotid stenosis) are eagerly awaited. Until then, clinicians should continue to weigh individual patient risks and benefits when referring patients for surgical treatment of carotid athero-sclerotic disease. Regardless of whether surgery is undertaken, maximal medical therapy with the use of antiplatelet agents, blood pressure control, and statin therapy remains the mainstay of treatment.

Whether carotid stenting has any advantage over carotid surgery (endarterectomy)—and for which patients—is still a topic of study and debate.

Treatment of carotid atherosclerosis and stenosis is important in preventing stroke and its comorbidities. Today, three main treatments exist: medical management (lipid-lowering, antihypertensive, and antiplatelet therapy), surgery, and, more recently, carotid angioplasty and stenting. The rationale for these treatments is to decrease the risk of cerebral infarction by stabilizing or removing plaque and improving blood flow.

Surgery has proven beneficial in patients with symptomatic carotid stenosis greater than 50% or asymptomatic stenosis greater than 60%, but it is risky in some patients. Stenting has evolved in part from the success of surgery and the need for alternative treatments for patients who are at unacceptable risk of perioperative complications. However, it does not have a clear advantage over surgery in patients at average risk. Further, its use in patients with asymptomatic stenosis of any severity is still controversial.

In this paper we review the major trials of carotid endarterectomy and stenting and summarize what we know today about who should undergo these therapies.

NOT ALL STROKES ARE DUE TO CAROTID ATHEROSCLEROSIS

Depending on the institution’s referral pattern and population served, between 80% and 90% of strokes are ischemic (the rest being hemorrhagic).¹ Atherosclerosis of large arteries (typically defined as more than 50% stenosis of a major brain artery or branch cortical artery²) is just one cause of ischemic stroke, but it is an important one. Other identifiable causes of ischemic stroke include cardioembolism and small-artery occlusion (lacunar stroke), and some cases are idiopathic.

Large-artery atherosclerotic disease can damage the brain gradually, with carotid stenosis resulting in hypoperfusion and subsequent cerebral infarction. More commonly, however, the carotid plaque often seen in large-artery atherosclerotic disease can ulcerate and occlude the vessel acutely or generate platelet aggregates that may embolize, resulting in cerebral infarction or transient ischemic attack.

In the Lausanne Stroke Registry,³ the rate of ischemic stroke in patients with a greater than 50% large-artery stenosis ranged from 27% in 1979 to 17% in 2003, the decline likely being due to therapeutic advances.

SURGERY BEATS MEDICAL THERAPY FOR CAROTID ATHEROSCLEROSIS

Four landmark trials provided substantial evidence that carotid endarterectomy is better than medical management in patients with symptomatic or asymptomatic high-grade stenosis. These trials indirectly paved the way for carotid stenting.

The North American Symptomatic Carotid Endarterectomy Trial (NASCET)

Patients at 50 clinical centers who had had a hemispheric or retinal transient ischemic attack or a nondisabling stroke were randomized to undergo surgery (carotid endarterectomy) or no surgery. All patients received maximal medical management consisting of blood pressure control, lipid management if indicated, and antiplatelet therapy with aspirin. At baseline, 37% of patients were taking 650 mg or more of aspirin per day, and 11% were taking less than 325 mg per day. The patients were stratified into two prespecified groups on the basis of the severity of carotid stenosis: those with narrowing of 30% to 69% and those with narrowing of 70% to 99%.

Results in high-grade stenosis. In August 1991, the investigators published their results in patients with symptomatic high-grade (70%–99%) stenosis.⁴ Surgical treatment was more beneficial than medical management alone: the cumulative risk of any ipsilateral stroke at 2 years was 26% in the medical group and 9% in the surgical group, an absolute risk reduction of 17%. The benefit of endarterectomy was still apparent at 8 years of follow-up.⁵

Results in moderate stenosis. In 1998, the investigators published their results in patients with symptomatic moderate (< 70%) stenosis.⁵ Surgery was more beneficial than medical therapy in this subgroup as well: at 5 years, the rate of any ipsilateral stroke in patients with 50% to 69% stenosis was 15.7% in those treated surgically and 22.2% in those treated medically (P = .045). In patients with less than 50% stenosis, the 5-year stroke rate was not significantly lower with endarterectomy than with medical therapy.

The European Carotid Surgery Trial (ECST)

The ECST,⁶ published in 1998, corroborated the NASCET findings. This multicenter, randomized, controlled trial enrolled 3,024 patients with symptoms of at least one transient ischemic attack in the distribution of one or both carotid arteries.

Results. In patients with stenosis of greater than 80% (60% by the NASCET criteria for calculating angiographic stenosis), the frequency of major stroke or death at 3 years was 26.5% in the control group and 14.9% in the surgery group, an absolute difference of 11.6%.

The Endarterectomy for Asymptomatic Carotid Artery Stenosis (ACAS) trial

The NASCET and ECST studies made it clear that select groups of patients with symptomatic carotid stenosis benefit from carotid endarterectomy. But what about patients with stenosis but no prior stroke?

The ACAS trial aimed to find out.⁷ In this pivotal study, 1,662 patients with asymptomatic carotid artery stenosis greater than 60% were randomized to receive either medical therapy alone or medical plus surgical therapy.

Results were published in 2004. After a median follow-up of 2.7 years, the aggregate 5-year risk of ipsilateral stroke, any perioperative stroke, or death was estimated to be 5.1% in the surgical group and 11.0% in the medical group, a relative risk reduction of 53%. However, for surgery to be beneficial, the rate of perioperative death and other serious complications had to be less than 3%, and the expected patient survival had to be at least 5 years.

Of note, the benefit of carotid endarterectomy in this study was predominantly in men, with less of a benefit for women and diabetic patients. Furthermore, even though endarterectomy was beneficial in this asymptomatic cohort, the overall benefit in terms of stroke risk reduction was small compared with that in NASCET and ECST, in which patients had symptomatic disease.

The Asymptomatic Carotid Surgery Trial (ACST)

In this European version of ACAS, published in 2004, 3,120 patients with asymptomatic carotid narrowing on ultrasonography were randomized to undergo surgery or medical therapy.

Results. The risk of stroke or death within 30 days of carotid endarterectomy was 3.1%. In patients younger than 75 years who had carotid narrowing of 70% or more, immediate surgery decreased the net 5-year stroke risk from 12% to 6%.⁸

WHO SHOULD NOT UNDERGO CAROTID ENDARTERECTOMY?

From these studies, we can conclude that patients with symptomatic carotid stenosis of 50% or greater and patients with asymptomatic stenosis of 60% or greater benefit from carotid endarterectomy, but only if the perioperative rate of death and other serious complications is less than 3%.⁷

What are the risk factors for complications during this surgery? In 2006, Cremonesi et al,⁹ in a consensus paper, defined patients as being at high risk if they had any of the following:

• Contralateral laryngeal nerve palsy
• Radiation therapy to the neck
• Previous carotid endarterectomy with recurrent stenosis
• Lesions high in the cervical internal carotid artery or below the clavicle in the common carotid artery
• Severe tandem lesions
• Age greater than 80 years
• Severe pulmonary disease
• Congestive heart failure (New York Heart
• Association class 3 or 4) or known severe left ventricular dysfunction
• Open heart surgery needed within 6 weeks
• Myocardial infarction within the past 4 weeks
• Unstable angina
• Contralateral carotid occlusion.

Could endovascular treatment be the answer for these patients at high risk who should not undergo carotid endarterectomy? Indeed, the procedure is being studied extensively and performed more frequently. We summarize the major studies below.

STUDIES OF CAROTID STENTING VS ENDARTERECTOMY

The Carotid and Vertebral Artery Transluminal Angioplasty Study (CAVATAS)

This study, published in 2001,¹⁰ was the first randomized, multicenter trial to compare the risks and benefits of endovascular treatment (angioplasty with or without stenting) of carotid and vertebral artery stenosis with those of conventional surgery.

To be included, patients had to have carotid artery stenosis (symptomatic or asymptomatic) that was suitable for either carotid endarterectomy or endovascular treatment. Patients were not grouped on the basis of the severity of their stenosis, but the mean stenosis in randomized patients was 86%.

A total of 504 patients were enrolled, of whom 251 were randomized to undergo endovascular treatment. Most patients in this group underwent angioplasty alone, but 26% also received stents because of suboptimal vessel dilatation or at the discretion of the intervening physician.

The primary end point was any disabling stroke or death. Secondary end points were any ipsilateral stroke lasting longer than 7 days and the combination of death or disabling ipsilateral stroke.

The results showed no significant difference between endovascular treatment and surgery in any of these end points at 3 years. However, the overall rates of procedural stroke and death were nearly double those seen in NASCET and ECST. The investigators could not determine the reason for this higher risk, but they hypothesized that CAVATAS included patients at higher risk.

The restenosis rate was higher in the endovascular therapy group (14%) than in the surgery group (4%; P < .001). On the other hand, the surgery group had a higher rate of minor complications, including cranial nerve palsies and neck hematomas.

Carotid Revascularization With Endarterectomy or Stenting Systems (CARESS)

This prospective, multicenter, phase 2 trial, published in 2003, compared the outcomes of standard carotid endarterectomy vs carotid artery stenting using distal embolic protection devices.¹¹ All the patients in this study had at least 50% symptomatic stenosis or 75% asymptomatic stenosis.

Results. At 30 days, 7 (2.4%) of 254 patients in the endarterectomy group had had strokes, and one of the 7 patients with stroke died, so the combined rate of stroke or death (the primary end point) was 2.4%. In the stenting group, 3 (2.1%) of 143 patients had strokes and no patients died. Overall, there was no significant difference in the composite of death, stroke, or myocardial infarction (the secondary end point): 3% for carotid endarterectomy and 2% for stenting patients.

The Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial

In this trial,¹² published in 2004, patients had to have either symptomatic carotid disease with 50% stenosis or greater or asymptomatic stenosis of 80% or greater, determined by ultrasonography. Further, all patients had to have at least one comorbid condition that increased their perioperative risk. Up until this point, no trial had strictly defined patients at increased risk for complications after carotid endarterectomy and assessed subsequent outcomes. The risk factors included severe cardiac or pulmonary disease, age greater than 80, postendarterectomy carotid stenosis, previous neck surgery, previous neck radiation, contralateral recurrent laryngeal nerve palsy, and contralateral carotid occlusion.

Patients were randomized to undergo carotid artery stenting with distal protection or carotid endarterectomy.

The primary end points of this study were the cumulative incidence of major cardiovascular events at 1 year; death, stroke, or myocardial infarction within 30 days of intervention; and ipsilateral stroke between 31 days and 1 year. Secondary outcomes measured were the rates of target-vessel recanalization at 1 year, cranial nerve palsy, and surgical site complications.

Results. The rate of stroke or death was similar in both groups. The stenting group had fewer adverse cardiac events (mainly non-Q-wave myocardial infarction) than the surgery group. At 1 year the rate of major ipsilateral stroke was 3.3% in the endarterectomy group vs 0% in the stenting group (the difference was not significant), and the cardiovascular event rates continued to be higher in the endarterectomy group.

The investigators noted that myocardial infarction was included as a primary end point because patients with atherosclerotic vascular disease who undergo either stenting or endarterectomy are at a substantial risk of myocardial infarction, and a Q-wave or a non-Q-wave myocardial infarction in the perioperative period increases the risk of future complications and death. A perioperative non-Q-wave infarction increases the risk of death by a factor of 6 and increases the risk of myocardial infarction by a factor of 27 in the subsequent 6 months.

Overall, this study presents evidence that stenting, using distal embolic protection devices, is not inferior to endarterectomy and has fewer cardiovascular complications in patients who have at least one risk factor.

This recent multicenter, randomized study¹³ was designed to determine if stenting is as good as (not inferior to) carotid endarterectomy in patients with symptomatic carotid stenosis of at least 60%. The primary end point was to be the incidence of stroke or death within 30 days after treatment. However, the trial was stopped early after the inclusion of 527 patients for reasons of safety and futility.

Results. The 30-day incidence of any stroke or death was higher in the stenting group (9.6% vs 3.9%). The relative risk of any stroke or death after stenting as compared with endarterectomy was 2.5. The 30-day incidence of disabling stroke or death was also higher in the stenting group (3.4% vs 1.5%; relative risk 2.2). At 6 months, the incidence of any stroke or death was 6.1% after endarterectomy and 11.7% after stenting (P = .02). There was a trend toward more major local complications after stenting and systemic complications after endarterectomy. Cranial-nerve injury was more common after endarterectomy than after stenting (as expected). Overall, death and stroke rates were lower at 1 month and 6 months with endarterectomy than with stenting.

The Stent-Protected Angioplasty Versus Carotid Endarterectomy (SPACE) trial

This randomized, multicenter study,¹⁴ published in 2006, was also designed to compare the safety and efficacy of carotid stenting and endarterectomy. Some 1,200 patients with symptomatic carotid artery stenosis confirmed by ultrasonography were randomly assigned within 180 days of a transient ischemic attack or moderate stroke to undergo carotid artery stenting (n = 605) or carotid endarterectomy (n = 595). The primary end point was ipsilateral ischemic stroke or death 30 days after the procedure. A total of 1,183 patients were included in the analysis.

Results. The rate of the primary end point was 6.84% with stenting and 6.34% with endarterectomy. The study failed to prove the noninferiority of carotid artery stenting compared with carotid endarterectomy for the periprocedural complication rate. Results at 6 to 24 months are awaited.

The Carotid Revascularization Endarterectomy Versus Stenting (CREST) trial

Perhaps the most anxiously awaited results are those of the CREST trial,¹⁵ funded by the National Institutes of Health. This is a prospective, randomized, parallel, two-arm, multicenter clinical trial with blinded end point evaluation. Anticipated enrollment will include 2,500 patients. Patients are eligible for enrollment if they have symptoms of carotid stenosis within 180 days of a stroke or transient ischemic attack with ipsilateral carotid stenosis of at least 50% by angiography (70% by ultrasonography), or if they have asymptomatic carotid stenosis of at least 60% by angiography (70% by ultrasonography).

Patients are being randomized to undergo either carotid artery stenting or carotid endarterectomy. All receive aspirin as anti-platelet therapy, treatment for hypertension, and management of other stroke risk factors. Follow-up will last 4 years, with clinic visits at 1, 6, 12, 18, 24, 30, 36, 42, and 48 months. Primary outcome measures will be rates of death, stroke, or myocardial infarction at 30 days postoperatively, and ipsilateral stroke at 30 days postoperatively.

As of February 2007, 1,506 patients had been enrolled and 1,453 had been randomized at 94 sites in North America.

MEDICAID AND MEDICARE NOW PAY FOR THESE THERAPIES

An important practical consideration for patients and physicians is whether Medicaid and Medicare will pay for these therapies.

In July 2001, Medicare began to cover percutaneous transluminal angioplasty of the carotid artery with concurrent stent placement, when furnished in accordance with US Food and Drug Administration (FDA) protocols governing Category B (nonexperimental) investigational device exemption clinical trials.¹⁶ Angioplasty of the carotid artery, when provided solely for the purpose of carotid artery dilation concurrent with carotid stent placement, is considered to be a reasonable and necessary service when provided in the context of clinical trials.

In March 2005, Medicare began to provide coverage for percutaneous transluminal angioplasty of the carotid artery concurrent with the placement of an FDA-approved carotid stent with embolic protection for the following groups of patients:

• Those who would be at high risk during carotid endarterectomy and who also have symptomatic carotid artery stenosis of 70% or greater. Coverage is limited to procedures performed using FDA-approved carotid artery stenting systems and embolic protection devices.
• Those who would be at high risk during endarterectomy and who have symptomatic carotid artery stenosis of 50% to 70%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting post-approval.
• Those who would be at high risk during carotid endarterectomy and have asymptomatic carotid artery stenosis greater than 80%, in accordance with the Category B Investigational Device Exemption clinical trials regulation, as a routine cost under the clinical trials policy, or in accordance with the national coverage determination on carotid artery stenting postapproval studies.

As noted above, Medicare and Medicaid will only cover carotid stenting if the stent system is FDA-approved, with concomitant use of a distal embolic protection device. However, in view of conflicting data from stenting trials to date, including EVA-3S¹³ and SPACE,¹⁴ it remains to be seen if emboli protection devices significantly reduce periprocedural stroke rates. The FDA recommends that if it is not technically possible to use one of these devices, then the procedure should be aborted due to safety issues.

These coverage decisions are an important practical aspect of carotid stenting and they should be familiar to physicians when they see and refer patients with carotid disease.

WHAT CAN WE SAY AT THIS POINT?

Given the multiple recent and ongoing trials of stenting vs endarterectomy in carotid stenosis, debate continues as to what the role of stenting will be in the future. What can we say at this point?

In patients with asymptomatic carotid stenosis of greater than 60% or symptomatic carotid stenosis of greater than 50%, carotid endarterectomy has been proven to be superior to medical therapy alone.

The efficacy and safety of carotid stenting compared with carotid endarterectomy is still uncertain. In the trials reviewed above, carotid stenting did not appear to have a clear advantage over endarterectomy in patients at average surgical risk. Stenting may be most advantageous when used in patients with symptomatic carotid stenosis who would be at high operative risk, as indicated by the SAPPHIRE trial.

In patients with severe but asymptomatic carotid stenosis who are at high operative risk, the addition of carotid angioplasty and stenting to maximum medical therapy remains controversial. The periprocedural complication rate in these patients may actually exceed the rate of stroke in asymptomatic patients with greater than 60% stenosis who do not undergo stenting or surgery. In addition, subgroup analyses of patients with 70% to 99% symptomatic stenosis in various trials show that surgical benefit is greater in men than in women, and it remains to be seen whether there is any benefit in women with moderate stenoses, asymptomatic lesions, or both.¹⁷

Further experience and study are needed, and the results of the Carotid Stenting vs Surgery of Severe Carotid Artery Disease and Stroke Prevention in Asymptomatic Patients (ACT I) study (comparing stenting and surgery in asymptomatic carotid stenosis), and the ongoing CREST trial (comparing stenting and surgery in symptomatic and asymptomatic carotid stenosis) are eagerly awaited. Until then, clinicians should continue to weigh individual patient risks and benefits when referring patients for surgical treatment of carotid athero-sclerotic disease. Regardless of whether surgery is undertaken, maximal medical therapy with the use of antiplatelet agents, blood pressure control, and statin therapy remains the mainstay of treatment.