Will ATP IV Spell the End of Cholesterol Targets?

With the long-awaited updated clinical guidelines for cholesterol testing and management in adults expected to be released this summer, some experts are hoping that the recommendations will abandon the central focus of previous guidelines: treating to low-density-lipoprotein cholesterol targets.

The Adult Treatment Panel IV (ATP IV) guidelines will be the first since ATP III was released in 2001, which, like ATP II, identified LDL cholesterol lowering as the primary goal of treatment and identified different targets for LDL-lowering drug treatment. An update of the ATP III guidelines released in 2004 also supported the ATP III treatment goal of LDL cholesterol below 100 mg/dL in people at high risk, but also recommended an LDL cholesterol goal of 70 mg/dL or lower for patients at very high CHD risk, based on clinical trial evidence that had become available.

No Evidence Base for Goals

As the release of the guidelines has gotten closer, there have been calls by some cardiologists that LDL targets be abandoned – most notably in an open letter to the ATP IV committee, published in January, in which Dr. Rodney Hayward of the University of Michigan and Dr. Harlan Krumholz of Yale University encouraged the committee to "abandon the paradigm of treating patients to LDL targets, a change that will better align ATP IV with current clinical evidence" (Circ. Cardiovasc. Qual. Outcomes 2012:2-5).

"The dogma that treating to target is based on clinical trial evidence belies the fact that no clinical trial has yet tested this strategy," they wrote, referring to evidence in clinical trials indicating "that the use of statins, and not treatment to target, can reduce risk."

A tailored treatment strategy with high-dose statins based on risk prevents more coronary events while treating fewer people, and will save more lives than approaches aimed at LDL targets, Dr. Krumholz said in an interview.

"What we know about statins is that they reliably reduce risk in populations regardless of their baseline cholesterol levels," and that the benefit of an intervention depends on a patient’s underlying risk and the degree to which that intervention can reduce the risk, said Dr. Krumholz, director of the Yale-New Haven (Conn.) Hospital Center for Outcomes Research and Evaluation. "They seem to be effective no matter what the baseline LDL level is. So you’re not necessarily treating high cholesterol levels. At any level of cholesterol, it appears that these drugs lower risk."

He emphasized that he was not debating whether cholesterol is an underlying mechanism of atherosclerotic coronary heart disease. Rather, he considers statins as more of "a risk reduction pill" than a cholesterol reduction pill "because it doesn’t depend on someone’s cholesterol for its relative reduction in risk, so the people who benefit the most are the highest-risk patients."

In addition, Dr. Krumholz said that there has never been a large clinical trial that has sought to titrate patients to targets and evaluate outcomes, "so when people are identifying targets that everyone should be treated to, they are extrapolating data from many trials and drawing lines, but those are based on speculation." He cited the ACCORD study, which evaluated outcomes of tight glycemic control and found that achieving hemoglobin A_1c below 7% with an aggressive approach increased mortality.

By focusing on targets, he said, "you are agnostic to the strategy to get you to those targets." For example, recommending treatment to reduce cholesterol to below 70 mg/dL or 100 mg/dL implies using whatever means possible to achieve that goal and to use that metric to measure success.

Lab Results Aren’t Outcomes

The danger in that strategy is that there are drugs that reduce LDL or raise HDL that "don’t necessarily achieve outcomes that you would expect for patients," Dr. Krumholz said. For example, fenofibrate is promoted as lowering triglyceride levels and raising HDL levels, but two large trials of fenofibrate’s impact on patient outcomes have been negative and there have been no positive trials, he pointed out.

It is not yet known whether treatment with ezetimibe, an inhibitor of intestinal cholesterol absorption that reduces LDL cholesterol, improves patient outcomes – which is being studied in an ongoing trial – and this uncertainty is reflected in the drug’s label. "So the focus on the target often obscures the fact that not all drugs that can get you to target have been shown to improve outcomes," he added said. (The ezetimibe label includes the statement that the effect of the drug on cardiovascular morbidity and mortality "has not been determined.")

Then there are the recent examples of the once highly promising cholesteryl ester transfer protein (CETP) inhibitors. Torcetrapib, a potent CETP inhibitor, was associated with extremely favorable changes in HDL cholesterol levels as well as reductions in triglycerides and LDL cholesterol in a study that was terminated early because treatment was associated with an increase in mortality and cardiac events; the drug was never approved. If the manufacturer had not conducted this study, the degree of harm identified in the study would have been extremely difficult to detect once it was approved and used in clinical practice, said Dr. Krumholz, the Harold H. Hines Jr. Professor of Medicine and professor of investigative medicine and of public health at Yale University.

This is another example of a drug "doing great things to the lab test" that did not translate into clinical benefit, he said. "What we’re learning is that manipulating risk factors is tricky business, and some drugs and strategies may change them but will not favorably change the risk."

He also referred to the development of another CETP inhibitor, dalcetrapib, which was halted after what the manufacturer, Roche, recently announced was a lack of clinically meaningful efficacy of the drug in a phase III trial of patients with stable coronary heart disease following an acute coronary syndrome.

The lack of evidence that the targeted approach improves patient outcomes, he cited, includes the AIM-HIGH study, which found no additional benefit of niacin when added to statin therapy in patients with atherosclerotic cardiovascular disease and LDL cholesterol levels less than 70 mg/dL, even though niacin was associated with significant improvements in HDL cholesterol and triglyceride levels (N. Engl. J. Med. 2011;365:2255-67).

Dr. Krumholz also cited the JUPITER study, which found that treatment with rosuvastatin was associated with a significant reduction in major cardiovascular events among people who had LDL cholesterol levels below 130 mg/dL and elevated C-reactive protein levels (N. Engl. J. Med. 2008;359:2195-2207).

"We want to know that the strategies we employ actually help people and not just their lab tests, and even though the science is elegant that underlies the lipid hypothesis and no one is disputing the relationships and studies that have been conducted, we are beginning to recognize these drugs have a lot of effects," he said. "We know in an open population that these factors will predict risk, but don’t know whether changing them and that different strategies of changing them all have same benefit associated with them."

Recent concerns about reports of hyperglycemia and cognitive effects associated with statins emphasize that, for patients with little to gain from treatment, there is "no reason" to be exposed to this risk, but for those who have a lot to benefit from treatment, the risk may be worth taking, he said. In February, the Food and Drug Administration announced that memory loss and confusion associated with statins reversed once the statin was stopped and that reports of hyperglycemia had been added to statin labels.

A patient’s choice should also be considered, and when discussing treatment with patients, Dr. Krumholz likes to give them a sense of how many people need to take a medication like a statin for one person to benefit. "Even for a number that seems small, like 25, which is good in medicine, a patient may choose not to take it," he said, adding that treatment "should be customized, both to the patient’s preferences and to their risk."

Individual Risk Matters

In an interview, Dr. Robert Vogel, clinical professor at the University of Colorado, Denver, said that he agrees that the "the basic concept of getting to an LDL goal is wrong," and while he would not abandon LDL cholesterol as a risk factor, "I would certainly throw it out as a target." He said that this view is shared by a majority of lipidologists but a minority of cardiologists overall.

Courtesy Oscar G. Johnson VA Medical Center

He also cited the JUPITER study as well as the Treating to New Targets (TNT) study, which, he pointed out, compared two different atorvastatin doses, not two different LDL goals. This is representative of what is actually known – the effects of larger versus smaller treatment doses, or treatment versus placebo. "We never have tested treatment to specific goals," so those effects are unknown, but that "is how the ATP guidelines have been written," he noted.

In the TNT study of about 10,000 people with clinical evidence of CHD and LDL cholesterol levels below 130 mg/dL, intensive lipid lowering with 80 mg of atorvastatin was associated with a reduction in the risk of major coronary events that was significantly greater than that associated with 10 mg of atorvastatin daily (N. Engl. J. Med. 2005;352:1425-35).

Dr. Vogel said that he hoped that in ATP IV the concept of treating to goal will be changed to treating high-risk patients with high-dose effective statins, if tolerated, and matching treatment not to the LDL cholesterol, level but to an individual’s risk "or to the situation that has been demonstrated to be effective in prospective trials." This approach would be a substantial difference from previous guidelines and would be "changing more towards something that looks like treating heart failure," for which treatment is not based on particular parameters but on treatments that have been demonstrated to be effective.

Calls Met With Silence

Along with his colleague Dr. Michael Miller of the University of Maryland, Dr. Vogel tried to bring the issue to the forefront 3 years ago by writing an editorial entitled "Cholesterol Goals: Moving from Numbers to Treatment" (Clin. Cardiol. 2009;32:106-8). And 3 years earlier, Dr. Hayward and his associates wrote a review article concluding that "there is clear and compelling evidence that most patients at high risk for cardiovascular disease should be taking at least a moderate dose of a statin if tolerated, even if their natural LDL cholesterol level is low. We could find no published high-quality clinical evidence supporting titration of lipid therapy based on proposed LDL cholesterol targets" (Ann. Intern. Med. 2006;145:520-30).

These publications, and Dr. Krumholz’s letter earlier this year, have been answered with silence in the scientific literature.

VA Takes Tailored Approach

The Veterans Health Administration (VHA) is not waiting for guidelines to make the move to tailored treatment; it has already adopted this approach. Over the past 18 months, the country’s largest integrated health care system providing care to more than 8 million patients a year, has been reviewing the evidence on the role of LDL targets in managing cardiovascular risk, "particularly in light of concerns that performance measures tied too closely to targets might create incentives for overzealous pharmacotherapy that does not improve overall health," said Dr. Joe Francis, head of performance measurement for the VHA.

The Department of Veterans Affairs (VA) "determined that the strongest evidence supports treatment with statins as the core objective, independent of LDL target, so we reconfigured our internal performance measures to align with this evidence," he said in an interview. At the VA, lipid management in patients with ischemic heart disease or diabetes mellitus is now considered appropriate if the patient is on at least a moderate dose of a statin drug or the LDL is under 100 mg/dL, he added.

This decision was in line with recommendations issued by the American Heart Association and American College of Cardiology in November 2011, Dr. John Rumsfeld, national director of cardiology for the VA system, pointed out. Both he and Dr. Francis said they hoped that ATP IV will also be consistent with this evidence-based approach. Dr. Rumsfeld emphasized that the VHA performance measures "are not meant to supplant clinical judgment for individual cases where more aggressive treatment may be warranted."

Whether the panel is also headed in this direction could not be determined, as neither the panel chair, Dr. Neil J. Stone of Northwestern University, Chicago, nor the NHLBI press office would comment on the content of the report before its release. ATP IV is being developed simultaneously with updated reports on hypertension guidelines (JNC 8) and obesity guidelines (Obesity 2), which also are expected to be available for public review and comment this year and will be part of an integrated cardiovascular risk reduction guideline, according to the statement.

Dr. Krumholz said that he could not speculate as to what the recommendations will be, "but if they stick to the evidence, then they will transition ... to a different approach." Because of the public comment period after they are released, there will be ample opportunity for discussing and vetting the guidelines before they are finalized. What he considers most important is that "they adhere to the evidence and recognize that a target-based approach could lead to treatments with medicines that are yet unproven with regard to their benefit for patients."

Dr. Vogel suggested that the long delay in the release of ATP IV is an indicator that the panel is incorporating emerging data and likely struggling with the very issues detailed in this article.

The new guidelines, which will be in a draft format available for public review and comment, are expected to be released this summer, according to the National Heart, Lung, and Blood Institute.

None of the experts who contributed to this article had relevant financial disclosures.

With the long-awaited updated clinical guidelines for cholesterol testing and management in adults expected to be released this summer, some experts are hoping that the recommendations will abandon the central focus of previous guidelines: treating to low-density-lipoprotein cholesterol targets.

The Adult Treatment Panel IV (ATP IV) guidelines will be the first since ATP III was released in 2001, which, like ATP II, identified LDL cholesterol lowering as the primary goal of treatment and identified different targets for LDL-lowering drug treatment. An update of the ATP III guidelines released in 2004 also supported the ATP III treatment goal of LDL cholesterol below 100 mg/dL in people at high risk, but also recommended an LDL cholesterol goal of 70 mg/dL or lower for patients at very high CHD risk, based on clinical trial evidence that had become available.

No Evidence Base for Goals

As the release of the guidelines has gotten closer, there have been calls by some cardiologists that LDL targets be abandoned – most notably in an open letter to the ATP IV committee, published in January, in which Dr. Rodney Hayward of the University of Michigan and Dr. Harlan Krumholz of Yale University encouraged the committee to "abandon the paradigm of treating patients to LDL targets, a change that will better align ATP IV with current clinical evidence" (Circ. Cardiovasc. Qual. Outcomes 2012:2-5).

"The dogma that treating to target is based on clinical trial evidence belies the fact that no clinical trial has yet tested this strategy," they wrote, referring to evidence in clinical trials indicating "that the use of statins, and not treatment to target, can reduce risk."

A tailored treatment strategy with high-dose statins based on risk prevents more coronary events while treating fewer people, and will save more lives than approaches aimed at LDL targets, Dr. Krumholz said in an interview.

"What we know about statins is that they reliably reduce risk in populations regardless of their baseline cholesterol levels," and that the benefit of an intervention depends on a patient’s underlying risk and the degree to which that intervention can reduce the risk, said Dr. Krumholz, director of the Yale-New Haven (Conn.) Hospital Center for Outcomes Research and Evaluation. "They seem to be effective no matter what the baseline LDL level is. So you’re not necessarily treating high cholesterol levels. At any level of cholesterol, it appears that these drugs lower risk."

He emphasized that he was not debating whether cholesterol is an underlying mechanism of atherosclerotic coronary heart disease. Rather, he considers statins as more of "a risk reduction pill" than a cholesterol reduction pill "because it doesn’t depend on someone’s cholesterol for its relative reduction in risk, so the people who benefit the most are the highest-risk patients."

In addition, Dr. Krumholz said that there has never been a large clinical trial that has sought to titrate patients to targets and evaluate outcomes, "so when people are identifying targets that everyone should be treated to, they are extrapolating data from many trials and drawing lines, but those are based on speculation." He cited the ACCORD study, which evaluated outcomes of tight glycemic control and found that achieving hemoglobin A_1c below 7% with an aggressive approach increased mortality.

By focusing on targets, he said, "you are agnostic to the strategy to get you to those targets." For example, recommending treatment to reduce cholesterol to below 70 mg/dL or 100 mg/dL implies using whatever means possible to achieve that goal and to use that metric to measure success.

Lab Results Aren’t Outcomes

The danger in that strategy is that there are drugs that reduce LDL or raise HDL that "don’t necessarily achieve outcomes that you would expect for patients," Dr. Krumholz said. For example, fenofibrate is promoted as lowering triglyceride levels and raising HDL levels, but two large trials of fenofibrate’s impact on patient outcomes have been negative and there have been no positive trials, he pointed out.

It is not yet known whether treatment with ezetimibe, an inhibitor of intestinal cholesterol absorption that reduces LDL cholesterol, improves patient outcomes – which is being studied in an ongoing trial – and this uncertainty is reflected in the drug’s label. "So the focus on the target often obscures the fact that not all drugs that can get you to target have been shown to improve outcomes," he added said. (The ezetimibe label includes the statement that the effect of the drug on cardiovascular morbidity and mortality "has not been determined.")

Then there are the recent examples of the once highly promising cholesteryl ester transfer protein (CETP) inhibitors. Torcetrapib, a potent CETP inhibitor, was associated with extremely favorable changes in HDL cholesterol levels as well as reductions in triglycerides and LDL cholesterol in a study that was terminated early because treatment was associated with an increase in mortality and cardiac events; the drug was never approved. If the manufacturer had not conducted this study, the degree of harm identified in the study would have been extremely difficult to detect once it was approved and used in clinical practice, said Dr. Krumholz, the Harold H. Hines Jr. Professor of Medicine and professor of investigative medicine and of public health at Yale University.

This is another example of a drug "doing great things to the lab test" that did not translate into clinical benefit, he said. "What we’re learning is that manipulating risk factors is tricky business, and some drugs and strategies may change them but will not favorably change the risk."

He also referred to the development of another CETP inhibitor, dalcetrapib, which was halted after what the manufacturer, Roche, recently announced was a lack of clinically meaningful efficacy of the drug in a phase III trial of patients with stable coronary heart disease following an acute coronary syndrome.

The lack of evidence that the targeted approach improves patient outcomes, he cited, includes the AIM-HIGH study, which found no additional benefit of niacin when added to statin therapy in patients with atherosclerotic cardiovascular disease and LDL cholesterol levels less than 70 mg/dL, even though niacin was associated with significant improvements in HDL cholesterol and triglyceride levels (N. Engl. J. Med. 2011;365:2255-67).

Dr. Krumholz also cited the JUPITER study, which found that treatment with rosuvastatin was associated with a significant reduction in major cardiovascular events among people who had LDL cholesterol levels below 130 mg/dL and elevated C-reactive protein levels (N. Engl. J. Med. 2008;359:2195-2207).

"We want to know that the strategies we employ actually help people and not just their lab tests, and even though the science is elegant that underlies the lipid hypothesis and no one is disputing the relationships and studies that have been conducted, we are beginning to recognize these drugs have a lot of effects," he said. "We know in an open population that these factors will predict risk, but don’t know whether changing them and that different strategies of changing them all have same benefit associated with them."

Recent concerns about reports of hyperglycemia and cognitive effects associated with statins emphasize that, for patients with little to gain from treatment, there is "no reason" to be exposed to this risk, but for those who have a lot to benefit from treatment, the risk may be worth taking, he said. In February, the Food and Drug Administration announced that memory loss and confusion associated with statins reversed once the statin was stopped and that reports of hyperglycemia had been added to statin labels.

A patient’s choice should also be considered, and when discussing treatment with patients, Dr. Krumholz likes to give them a sense of how many people need to take a medication like a statin for one person to benefit. "Even for a number that seems small, like 25, which is good in medicine, a patient may choose not to take it," he said, adding that treatment "should be customized, both to the patient’s preferences and to their risk."

Individual Risk Matters

In an interview, Dr. Robert Vogel, clinical professor at the University of Colorado, Denver, said that he agrees that the "the basic concept of getting to an LDL goal is wrong," and while he would not abandon LDL cholesterol as a risk factor, "I would certainly throw it out as a target." He said that this view is shared by a majority of lipidologists but a minority of cardiologists overall.

Courtesy Oscar G. Johnson VA Medical Center

He also cited the JUPITER study as well as the Treating to New Targets (TNT) study, which, he pointed out, compared two different atorvastatin doses, not two different LDL goals. This is representative of what is actually known – the effects of larger versus smaller treatment doses, or treatment versus placebo. "We never have tested treatment to specific goals," so those effects are unknown, but that "is how the ATP guidelines have been written," he noted.

In the TNT study of about 10,000 people with clinical evidence of CHD and LDL cholesterol levels below 130 mg/dL, intensive lipid lowering with 80 mg of atorvastatin was associated with a reduction in the risk of major coronary events that was significantly greater than that associated with 10 mg of atorvastatin daily (N. Engl. J. Med. 2005;352:1425-35).

Dr. Vogel said that he hoped that in ATP IV the concept of treating to goal will be changed to treating high-risk patients with high-dose effective statins, if tolerated, and matching treatment not to the LDL cholesterol, level but to an individual’s risk "or to the situation that has been demonstrated to be effective in prospective trials." This approach would be a substantial difference from previous guidelines and would be "changing more towards something that looks like treating heart failure," for which treatment is not based on particular parameters but on treatments that have been demonstrated to be effective.

Calls Met With Silence

Along with his colleague Dr. Michael Miller of the University of Maryland, Dr. Vogel tried to bring the issue to the forefront 3 years ago by writing an editorial entitled "Cholesterol Goals: Moving from Numbers to Treatment" (Clin. Cardiol. 2009;32:106-8). And 3 years earlier, Dr. Hayward and his associates wrote a review article concluding that "there is clear and compelling evidence that most patients at high risk for cardiovascular disease should be taking at least a moderate dose of a statin if tolerated, even if their natural LDL cholesterol level is low. We could find no published high-quality clinical evidence supporting titration of lipid therapy based on proposed LDL cholesterol targets" (Ann. Intern. Med. 2006;145:520-30).

These publications, and Dr. Krumholz’s letter earlier this year, have been answered with silence in the scientific literature.

VA Takes Tailored Approach

The Veterans Health Administration (VHA) is not waiting for guidelines to make the move to tailored treatment; it has already adopted this approach. Over the past 18 months, the country’s largest integrated health care system providing care to more than 8 million patients a year, has been reviewing the evidence on the role of LDL targets in managing cardiovascular risk, "particularly in light of concerns that performance measures tied too closely to targets might create incentives for overzealous pharmacotherapy that does not improve overall health," said Dr. Joe Francis, head of performance measurement for the VHA.

The Department of Veterans Affairs (VA) "determined that the strongest evidence supports treatment with statins as the core objective, independent of LDL target, so we reconfigured our internal performance measures to align with this evidence," he said in an interview. At the VA, lipid management in patients with ischemic heart disease or diabetes mellitus is now considered appropriate if the patient is on at least a moderate dose of a statin drug or the LDL is under 100 mg/dL, he added.

This decision was in line with recommendations issued by the American Heart Association and American College of Cardiology in November 2011, Dr. John Rumsfeld, national director of cardiology for the VA system, pointed out. Both he and Dr. Francis said they hoped that ATP IV will also be consistent with this evidence-based approach. Dr. Rumsfeld emphasized that the VHA performance measures "are not meant to supplant clinical judgment for individual cases where more aggressive treatment may be warranted."

Whether the panel is also headed in this direction could not be determined, as neither the panel chair, Dr. Neil J. Stone of Northwestern University, Chicago, nor the NHLBI press office would comment on the content of the report before its release. ATP IV is being developed simultaneously with updated reports on hypertension guidelines (JNC 8) and obesity guidelines (Obesity 2), which also are expected to be available for public review and comment this year and will be part of an integrated cardiovascular risk reduction guideline, according to the statement.

Dr. Krumholz said that he could not speculate as to what the recommendations will be, "but if they stick to the evidence, then they will transition ... to a different approach." Because of the public comment period after they are released, there will be ample opportunity for discussing and vetting the guidelines before they are finalized. What he considers most important is that "they adhere to the evidence and recognize that a target-based approach could lead to treatments with medicines that are yet unproven with regard to their benefit for patients."

Dr. Vogel suggested that the long delay in the release of ATP IV is an indicator that the panel is incorporating emerging data and likely struggling with the very issues detailed in this article.

The new guidelines, which will be in a draft format available for public review and comment, are expected to be released this summer, according to the National Heart, Lung, and Blood Institute.

None of the experts who contributed to this article had relevant financial disclosures.

With the long-awaited updated clinical guidelines for cholesterol testing and management in adults expected to be released this summer, some experts are hoping that the recommendations will abandon the central focus of previous guidelines: treating to low-density-lipoprotein cholesterol targets.

The Adult Treatment Panel IV (ATP IV) guidelines will be the first since ATP III was released in 2001, which, like ATP II, identified LDL cholesterol lowering as the primary goal of treatment and identified different targets for LDL-lowering drug treatment. An update of the ATP III guidelines released in 2004 also supported the ATP III treatment goal of LDL cholesterol below 100 mg/dL in people at high risk, but also recommended an LDL cholesterol goal of 70 mg/dL or lower for patients at very high CHD risk, based on clinical trial evidence that had become available.

No Evidence Base for Goals

As the release of the guidelines has gotten closer, there have been calls by some cardiologists that LDL targets be abandoned – most notably in an open letter to the ATP IV committee, published in January, in which Dr. Rodney Hayward of the University of Michigan and Dr. Harlan Krumholz of Yale University encouraged the committee to "abandon the paradigm of treating patients to LDL targets, a change that will better align ATP IV with current clinical evidence" (Circ. Cardiovasc. Qual. Outcomes 2012:2-5).

"The dogma that treating to target is based on clinical trial evidence belies the fact that no clinical trial has yet tested this strategy," they wrote, referring to evidence in clinical trials indicating "that the use of statins, and not treatment to target, can reduce risk."

A tailored treatment strategy with high-dose statins based on risk prevents more coronary events while treating fewer people, and will save more lives than approaches aimed at LDL targets, Dr. Krumholz said in an interview.

"What we know about statins is that they reliably reduce risk in populations regardless of their baseline cholesterol levels," and that the benefit of an intervention depends on a patient’s underlying risk and the degree to which that intervention can reduce the risk, said Dr. Krumholz, director of the Yale-New Haven (Conn.) Hospital Center for Outcomes Research and Evaluation. "They seem to be effective no matter what the baseline LDL level is. So you’re not necessarily treating high cholesterol levels. At any level of cholesterol, it appears that these drugs lower risk."

He emphasized that he was not debating whether cholesterol is an underlying mechanism of atherosclerotic coronary heart disease. Rather, he considers statins as more of "a risk reduction pill" than a cholesterol reduction pill "because it doesn’t depend on someone’s cholesterol for its relative reduction in risk, so the people who benefit the most are the highest-risk patients."

In addition, Dr. Krumholz said that there has never been a large clinical trial that has sought to titrate patients to targets and evaluate outcomes, "so when people are identifying targets that everyone should be treated to, they are extrapolating data from many trials and drawing lines, but those are based on speculation." He cited the ACCORD study, which evaluated outcomes of tight glycemic control and found that achieving hemoglobin A_1c below 7% with an aggressive approach increased mortality.

By focusing on targets, he said, "you are agnostic to the strategy to get you to those targets." For example, recommending treatment to reduce cholesterol to below 70 mg/dL or 100 mg/dL implies using whatever means possible to achieve that goal and to use that metric to measure success.

Lab Results Aren’t Outcomes

The danger in that strategy is that there are drugs that reduce LDL or raise HDL that "don’t necessarily achieve outcomes that you would expect for patients," Dr. Krumholz said. For example, fenofibrate is promoted as lowering triglyceride levels and raising HDL levels, but two large trials of fenofibrate’s impact on patient outcomes have been negative and there have been no positive trials, he pointed out.

It is not yet known whether treatment with ezetimibe, an inhibitor of intestinal cholesterol absorption that reduces LDL cholesterol, improves patient outcomes – which is being studied in an ongoing trial – and this uncertainty is reflected in the drug’s label. "So the focus on the target often obscures the fact that not all drugs that can get you to target have been shown to improve outcomes," he added said. (The ezetimibe label includes the statement that the effect of the drug on cardiovascular morbidity and mortality "has not been determined.")

Then there are the recent examples of the once highly promising cholesteryl ester transfer protein (CETP) inhibitors. Torcetrapib, a potent CETP inhibitor, was associated with extremely favorable changes in HDL cholesterol levels as well as reductions in triglycerides and LDL cholesterol in a study that was terminated early because treatment was associated with an increase in mortality and cardiac events; the drug was never approved. If the manufacturer had not conducted this study, the degree of harm identified in the study would have been extremely difficult to detect once it was approved and used in clinical practice, said Dr. Krumholz, the Harold H. Hines Jr. Professor of Medicine and professor of investigative medicine and of public health at Yale University.

This is another example of a drug "doing great things to the lab test" that did not translate into clinical benefit, he said. "What we’re learning is that manipulating risk factors is tricky business, and some drugs and strategies may change them but will not favorably change the risk."

He also referred to the development of another CETP inhibitor, dalcetrapib, which was halted after what the manufacturer, Roche, recently announced was a lack of clinically meaningful efficacy of the drug in a phase III trial of patients with stable coronary heart disease following an acute coronary syndrome.

The lack of evidence that the targeted approach improves patient outcomes, he cited, includes the AIM-HIGH study, which found no additional benefit of niacin when added to statin therapy in patients with atherosclerotic cardiovascular disease and LDL cholesterol levels less than 70 mg/dL, even though niacin was associated with significant improvements in HDL cholesterol and triglyceride levels (N. Engl. J. Med. 2011;365:2255-67).

Dr. Krumholz also cited the JUPITER study, which found that treatment with rosuvastatin was associated with a significant reduction in major cardiovascular events among people who had LDL cholesterol levels below 130 mg/dL and elevated C-reactive protein levels (N. Engl. J. Med. 2008;359:2195-2207).

"We want to know that the strategies we employ actually help people and not just their lab tests, and even though the science is elegant that underlies the lipid hypothesis and no one is disputing the relationships and studies that have been conducted, we are beginning to recognize these drugs have a lot of effects," he said. "We know in an open population that these factors will predict risk, but don’t know whether changing them and that different strategies of changing them all have same benefit associated with them."

Recent concerns about reports of hyperglycemia and cognitive effects associated with statins emphasize that, for patients with little to gain from treatment, there is "no reason" to be exposed to this risk, but for those who have a lot to benefit from treatment, the risk may be worth taking, he said. In February, the Food and Drug Administration announced that memory loss and confusion associated with statins reversed once the statin was stopped and that reports of hyperglycemia had been added to statin labels.

A patient’s choice should also be considered, and when discussing treatment with patients, Dr. Krumholz likes to give them a sense of how many people need to take a medication like a statin for one person to benefit. "Even for a number that seems small, like 25, which is good in medicine, a patient may choose not to take it," he said, adding that treatment "should be customized, both to the patient’s preferences and to their risk."

Individual Risk Matters

In an interview, Dr. Robert Vogel, clinical professor at the University of Colorado, Denver, said that he agrees that the "the basic concept of getting to an LDL goal is wrong," and while he would not abandon LDL cholesterol as a risk factor, "I would certainly throw it out as a target." He said that this view is shared by a majority of lipidologists but a minority of cardiologists overall.

Courtesy Oscar G. Johnson VA Medical Center

He also cited the JUPITER study as well as the Treating to New Targets (TNT) study, which, he pointed out, compared two different atorvastatin doses, not two different LDL goals. This is representative of what is actually known – the effects of larger versus smaller treatment doses, or treatment versus placebo. "We never have tested treatment to specific goals," so those effects are unknown, but that "is how the ATP guidelines have been written," he noted.

In the TNT study of about 10,000 people with clinical evidence of CHD and LDL cholesterol levels below 130 mg/dL, intensive lipid lowering with 80 mg of atorvastatin was associated with a reduction in the risk of major coronary events that was significantly greater than that associated with 10 mg of atorvastatin daily (N. Engl. J. Med. 2005;352:1425-35).

Dr. Vogel said that he hoped that in ATP IV the concept of treating to goal will be changed to treating high-risk patients with high-dose effective statins, if tolerated, and matching treatment not to the LDL cholesterol, level but to an individual’s risk "or to the situation that has been demonstrated to be effective in prospective trials." This approach would be a substantial difference from previous guidelines and would be "changing more towards something that looks like treating heart failure," for which treatment is not based on particular parameters but on treatments that have been demonstrated to be effective.

Calls Met With Silence

Along with his colleague Dr. Michael Miller of the University of Maryland, Dr. Vogel tried to bring the issue to the forefront 3 years ago by writing an editorial entitled "Cholesterol Goals: Moving from Numbers to Treatment" (Clin. Cardiol. 2009;32:106-8). And 3 years earlier, Dr. Hayward and his associates wrote a review article concluding that "there is clear and compelling evidence that most patients at high risk for cardiovascular disease should be taking at least a moderate dose of a statin if tolerated, even if their natural LDL cholesterol level is low. We could find no published high-quality clinical evidence supporting titration of lipid therapy based on proposed LDL cholesterol targets" (Ann. Intern. Med. 2006;145:520-30).

These publications, and Dr. Krumholz’s letter earlier this year, have been answered with silence in the scientific literature.

VA Takes Tailored Approach

The Veterans Health Administration (VHA) is not waiting for guidelines to make the move to tailored treatment; it has already adopted this approach. Over the past 18 months, the country’s largest integrated health care system providing care to more than 8 million patients a year, has been reviewing the evidence on the role of LDL targets in managing cardiovascular risk, "particularly in light of concerns that performance measures tied too closely to targets might create incentives for overzealous pharmacotherapy that does not improve overall health," said Dr. Joe Francis, head of performance measurement for the VHA.

The Department of Veterans Affairs (VA) "determined that the strongest evidence supports treatment with statins as the core objective, independent of LDL target, so we reconfigured our internal performance measures to align with this evidence," he said in an interview. At the VA, lipid management in patients with ischemic heart disease or diabetes mellitus is now considered appropriate if the patient is on at least a moderate dose of a statin drug or the LDL is under 100 mg/dL, he added.

This decision was in line with recommendations issued by the American Heart Association and American College of Cardiology in November 2011, Dr. John Rumsfeld, national director of cardiology for the VA system, pointed out. Both he and Dr. Francis said they hoped that ATP IV will also be consistent with this evidence-based approach. Dr. Rumsfeld emphasized that the VHA performance measures "are not meant to supplant clinical judgment for individual cases where more aggressive treatment may be warranted."

Whether the panel is also headed in this direction could not be determined, as neither the panel chair, Dr. Neil J. Stone of Northwestern University, Chicago, nor the NHLBI press office would comment on the content of the report before its release. ATP IV is being developed simultaneously with updated reports on hypertension guidelines (JNC 8) and obesity guidelines (Obesity 2), which also are expected to be available for public review and comment this year and will be part of an integrated cardiovascular risk reduction guideline, according to the statement.

Dr. Krumholz said that he could not speculate as to what the recommendations will be, "but if they stick to the evidence, then they will transition ... to a different approach." Because of the public comment period after they are released, there will be ample opportunity for discussing and vetting the guidelines before they are finalized. What he considers most important is that "they adhere to the evidence and recognize that a target-based approach could lead to treatments with medicines that are yet unproven with regard to their benefit for patients."

Dr. Vogel suggested that the long delay in the release of ATP IV is an indicator that the panel is incorporating emerging data and likely struggling with the very issues detailed in this article.

The new guidelines, which will be in a draft format available for public review and comment, are expected to be released this summer, according to the National Heart, Lung, and Blood Institute.

None of the experts who contributed to this article had relevant financial disclosures.