Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

JOURNAL CLUB 2009 - Cardiology

February 2009

Editorial Overview:

Jacques Genest, Jr., MD, FRCPC

Professor of Medicine, McGill University Montreal, Quebec

A large multinational trial has confirmed the hypothesis that high-sensitivity C-reactive protein (hsCRP) can identify patients at increased cardiovascular (CV) risk independent of LDL-C. The primary prevention study, called JUPITER (Justification of the Use of Statins in Prevention: an Intervention Trial Evaluating Rosuvastatin), demonstrated that when apparently healthy individuals with elevated hsCRP but without hyperlipidemia are treated with an aggressive lipidlowering agent, the risk of CV events can be reduced by a highly significant degree. The relative benefit was achieved independent of any other risk factors. The results of JUPITER are likely to alter current guidelines for risk stratification. Although the measurement of hsCRP is not likely to improve risk stratification in patients already at high risk or to be a useful screening tool in low-risk populations, it does appear to be useful for better defining those at intermediate risk.

In Canada as in many other industrialized countries, about half of myocardial infarctions (MIs), strokes, and other major CV events occur in patients without elevated LDL-C. The JUPITER trial, presented at the annual scientific sessions of the American Heart Association (AHA) and subsequently published (Ridker et al. N Engl J Med 2008;359:2195-207), has provided compelling evidence that hsCRP can identify a proportion of this population. JUPITER enrolled men over the age of 50 and women over the age of 60 who had elevated hsCRP but not hyperlipidemia. In this population, the lipid-lowering agent rosuvastatin reduced the risk of the primary composite end point of MI, stroke, arterial revascularization, hospitalization for unstable angina, or death from CV causes by 44% (HR 0.56; 95% CI, 0.46–0.69; P<0.00001) despite early termination after 1.9 years of follow-up. The robust risk reduction was consistent across all risk stratifications evaluated.

The size of the relative benefit from a primary prevention study that enrolled patients only on the basis of an elevated hsCRP has validated the hypothesis that hsCRP is a viable risk marker independent of LDL-C. This study was not designed to determine whether hsCRP is a treatable risk factor. Rather, the study design confirmed the hypothesis that apparently healthy older men and women without a prior history of CV disease and without hyperlipidemia can reduce their risk of a major CV event when treated with rosuvastatin, which produced large reductions in both LDL-C (50%) and hsCRP (37%) at a relatively modest once-daily dose of 20 mg.

JUPITER was a major clinical trial that enrolled 17,802 individuals in 26 countries including Canada. The primary inclusion criterion was an hsCRP =2 mg/L. The primary exclusion criteria were a LDL-C =3.4 mmol/L and any history or evidence of CV or cerebrovascular disease. Additional exclusion criteria included previous or current use of lipid-lowering therapy, current use of hormone replacement therapy, uncontrolled hypertension, diabetes, or any evidence of significant hepatic or renal dysfunction. The goal was to isolate a population for which hsCRP was the major risk variable. Study subjects were randomized to rosuvastatin 20 mg or placebo once daily.

Figure 1.

The Data Safety Monitoring Committee recommended termination of the study less than two years into a planned follow-up of five years because of the large advantage in the active treatment arm. At the time of termination, the trial was fully enrolled. When the full analyses were completed, the advantage for rosuvastatin included very large and highly statistically significant reductions in the major end points, such as MI (54%; P=0.0002) and stroke (48%; P=0.002), whether assessed collectively or individually. Despite the relatively brief follow-up, rosuvastatin was 4 associated with a 20% reduction (HR 0.80; 95% CI, 0.67–0.97; P=0.02) in death from any cause. Furthermore, the relative advantage of rosuvastatin was remarkably consistent across patient groups stratified by gender, age (over 65 vs. 65 or younger), smoking status, ethnic group, presence or absence of hypertension, presence or absence of family history of CV disease, body mass index, or presence or absence of metabolic syndrome.

Reinforcing the independent value of hsCRP, the relative risk reduction for a major CV event on rosuvastatin was independent of Framingham risk calculation (<10% vs. >10% over 10 years). In a subgroup of patients with no risk factors other than age for a CV event, the relative benefit of rosuvastatin over placebo was of a similar magnitude as that observed in the whole population.

The relative reductions in LDL-C and hsCRP in the active treatment arm were larger than those observed in previous statin trials. On the 20-mg dose of rosuvastatin, the median LDL-C was reduced to 1.4 mmol/L (55 mg/dL) with about 25% of patients achieving an LDL-C <1.0 mmol/L. The hsCRP levels were reduced from a median of 4.4 mg/L to approximately 2.0 mg/L. Although reductions in LDL-C and hsCRP are not interdependent, rosuvastatin is highly effective even at relatively modest doses for achieving large reductions in both. Similar reductions would be difficult to achieve in most patients even at the highest dos
tins.

Table 1.

<img2796|center>

These reductions in LDL-C and hsCRP in an apparently healthy population were well tolerated and safe. In a large study capable of revealing relatively rare events, there was no adverse safety signal of any kind. This included no significant difference or trend for differences between treatment groups for muscle weakness, myopathy or rhabdomyolysis. As reported in many previous statin trials, there was a small increase in the hemoglobin A1c (5.9% vs. 5.8%; P=0.001) and in physician-reported new cases of diabetes (3.0% vs. 2.4%; P=0.01) among those taking the statin. The rates of newly diagnosed cancer and death from cancer were both lower in the treatment arm.

Although the median follow-up was short, the large study population does provide reassurance about the safety of the study agent and of achieving very low LDL-C levels in the absence of prior evidence of vascular disease. In clinical trials overall, statins have proven to be remarkably well tolerated and safe. JUPITER builds on this evidence of safety by demonstrating that LDL-C reductions of 50% to the lowest mean LDL-C levels yet achieved in a large trial extends benefits and does not introduce risks. JUPITER contributes strongly to the premise that, beyond current targets, lower is better in regard to LDL-C. While the contribution of hsCRP reductions to the results of JUPITER is unclear, the reductions in LDL-C observed in JUPITER was obtained with only 20 mg of rosuvastatin. The goal for LDL-C, rather than a specific target, may simply be the lowest level achievable.

The evidence that hsCRP identifies patients who can benefit from statin therapy independent of LDL-C was predicted by earlier studies, such as AFCAPS/TexCAPS (Air Force/ Texas Coronary Atherosclerosis Prevention Study), but the proof of benefit now requires interpretation for its relevance to clinical practice. Consensus panels are being convened or will soon be convened in many countries, including Canada, to provide guidance on how these results should change risk stratification and treatment strategies.

The economics of healthcare will be a factor in the discussion. The issues not only include the increased cost of medicine for an expanded group of candidates for statins but the costs of screening and other associated expenses. However, hsCRP screening also offers an important opportunity to reduce the morbidity and mortality of CV disease and is likely to be warranted by
cost efficacy in many patients not currently treated.

Table 2.

<img2797|center>

In modifying the guidelines, it will be important to maintain the discussion within the confines of the current evidence. JUPITER demonstrates that identifying elevated hsCRP in men over the age of 50 and women over the age of 60 without hyperlipidemia yields a previously unidentified pool of patients who will benefit from intensive lipid lowering. These data do not support universal hsCRP screening in all individuals or even in all older individuals. Screening is not indicated or necessary in those who have already been identified as being at high risk whether due to hyperlipidemia, the presence of diabetes, or the presence of existing CV disease. The strategy of screening hsCRP in younger patients has not been tested and its benefits are therefore unknown. However, it is unlikely to be costeffective in those who are otherwise at low risk from the societal perspective.

The JUPITER study is a landmark trial that may very well change clinical practice. It demonstrates that hsCRP identifies a group of older patients without hyperlipidemia who can benefit from intensive lipid lowering with rosuvastatin. In younger patients, an elevated hsCRP may encourage treatment in a patient at borderline risk, although prospective studies are needed to confirm this strategy. We do not know from JUPITER whether hsCRP is a treatable risk factor, but we do know that patients with an elevated hsCRP achieve highly significant reductions in CV risk from a dose of rosuvastatin that produces reductions in both LDL-C and hsCRP.

Summary

Results of the JUPITER study demonstrate that an elevated plasma hsCRP level identifies a group of older individuals who can reduce their CV risk with 20 mg rosuvastatin even in the absence of hyperlipidemia or CV disease. The risk reductions were achieved at much lower median LDL-C levels than are currently recommended as targets even in high-risk patients. This provides a new avenue for risk management. New guidelines based on these findings are anticipated. The study does not encourage indiscriminate screening for hsCRP, but it does suggest that a substantial proportion of patients at high risk of CV disease who are missed by current risk factor screening may be identified through hsCRP.

questions and answers

Panel

Beth Abramson, MD, MSc, FRCPC, FACC, University of Toronto

Todd Anderson, MD, FRCPC, University of Calgary

Jean Davignon, OC, MD, MSc, FRCPC, FACP, FAHA, FRSC, Université de Montréal

Stephen Fort, MB ChB, MRCP, FRCPC, Dalhousie University

Jacques Genest, Jr., MD, FRCPC, McGill University

Ruth McPherson, MD, PhD, FRCPC, University of Ottawa

Subodh Verma, MD, PhD, FRCSC, FAHA, University of Toronto

What are the clinical implications of the JUPITER results?

Dr. Davignon: Essentially, JUPITER results emphasize the importance of statin therapy in primary prevention of CAD in patients whose plasma LDL-C is not at levels considered at present as a major risk determinant. They bring forth a new argument for the importance in humans of chronic inflammatory processes in atherogenesis as a CVD risk determinant. They also reveal dramatically a heretofore neglected cohort of patients who are at risk primarily because of an inflammatory driving force consistent with, and expected from, animal experimentation. They also enhance confidence in the efficacy and safety of a 20-mg dose of rosuvastatin for CVD prevention and treatment. The results come at an opportune time when the value of lowering cholesterol alone is being challenged.

Dr. Verma: JUPITER is a truly revolutionary trial in primary prevention. JUPITER has closed the loop on the inflammation hypothesis as it relates to CV disease. Inflammation is a target of statin therapy and results in robust reductions in death, MI, stroke and revascularization in patients who were previously not candidates for statin therapy. Levels of hsCRP need to be measured in patients in the low and intermediate risk categories, and if >2 mg/L statin therapy should be considered, irrespective of their LDL-C level. Dr. McPherson: These results are clinically relevant in deciding which individuals without clinical evidence of CV disease are likely to benefit most from aggressive lipid lowering. The results of JUPITER take us beyond the paradigm of Framingham by providing an additional tool to evaluate risk beyond conventional risk factors. This is important because we know that a substantial proportion of first events occur in patients without overt hyperlipidemia.

Dr. Fort: The results suggest that there is a group of individuals identified with hsCRP in whom statin therapy is not currently indicated, but who will benefit from statin therapy. The implication is that hsCRP is extremely useful for classifying and recognizing CV risk and it is likely to change clinical practice.

Dr. Anderson: JUPITER extends our ability to assess risk over what we are currently doing. This is the first effort to use a model that really falls outside the sphere of Framingham. The protection from events provided by aggressive lipid lowering was highly statistically significant in those with elevated hsCRP, but it is important to recognize that the absolute event rates were low. It still seems likely that there will be an acceptable cost efficacy, but we need to see this.

Dr. Abramson: This is an exciting trial that suggests we can evaluate risk independent of lipids and other conventional factors. The results of JUPITER also speak to the underlying pathology by providing strong support for the hypothesis that inflammation drives events.

Dr. Genest: JUPITER indicates that hsCRP, along with other risk factors, such as hypertension, hypercholesterolemia, diabetes and smoking, is useful for evaluating the likelihood of a CV event. We have had other evidence that hsCRP is a marker of risk, but JUPITER provides guidance for how it might be used in clinical practice.

Do you expect JUPITER results to alter cholesterol treatment guidelines?

Dr. Genest: Yes, I think changes in guidelines are coming and that these are important for putting the routine measurement of hsCRP into context with public health issues, such as cost. I think we need Canadian guidelines to explain how the results of JUPITER should translate into clinical practice.

Dr. McPherson: These data are likely to influence clinical lipid guidelines. In Canada, the 2006 guidelines did indicate that hsCRP could be considered for risk assessment, but JUPITER has substantially increased the confidence in hsCRP as a marker of risk and in defining a population that will benefit most from high-dose statin therapy. This may be especially helpful in patients whose risk is considered borderline by other measures. An elevated hsCRP may provide additional support for therapy. However, we need to think carefully about the age and Framingham risk levels where hsCRP will be most cost-effective and provide information not otherwise available.

Dr. Fort: Currently, the major factors for stratifying risk are age, lipid levels, blood pressure and smoking status, although these should be influenced by other considerations, such as family history of CV disease or the presence of metabolic syndrome. In those judged to be at high risk of vascular disease, statin therapy is appropriate and we do not need hsCRP. In those with moderate or lower risk, and perhaps older individuals, JUPITER says that hsCRP can identify a larger group who will benefit from statins. I expect the guidelines to reflect these findings.

Dr. Anderson: The potential role for hsCRP measurements is in the intermediate-risk patient, but I would like to see more cost analyses to confirm that the risk reductions achieved in JUPITER can be reproduced cost-effectively in routine practice. It is likely that such analyses are coming, and this will be important for rational guidelines.

Dr. Verma: Yes, hsCRP measurement to identify patients who should be candidates for statin therapy is likely to be a strong recommendation in the guidelines worldwide.

Dr. Abramson: I suspect that the results will be incorporated into new guidelines to some level, but I think it is necessary to be cautious about generalizing the importance of hsCRP outside of the population studied in JUPITER. These patients were already at increased risk because of their age. I do not think we can draw any conclusions from JUPITER about the role of evaluating hsCRP on younger patients.

Dr. Davignon: Yes. Guidelines should recognize the paradigm shift JUPITER results bring about and deal with it. The focus should be on: 1) how to identify those subjects who appear to be at high risk but cannot be readily detected by the current risk assessment measures. Not a simple task, as we have little information from JUPITER yet as to what characterizes those, among the 6375 subjects with an hsCRP over 2 mg/L, who had an event in the absence of other classical risk factors. It is not obvious whether family history of CVD helps detect them from the data published, as only 2045 subjects had a family history of CAD in the entire cohort; what is the absolute risk reduction in this subgroup, what is the NNT? 2) how to optimize hsCRP measurements in practice so that they bring to treatment those who will benefit maximally, provide the highest yield of positive hits with as few false-negatives as possible, and contain cost; 3) establishing to what extent hsCRP measurements are available and standardized across the country, arriving at a consensus regarding cut-off levels in the Canadian population, warning on sources of variation for decision to measure and for interpretation of results; 4) relating hsCRP levels to presence of other CAD risk factors for risk assessment. It is my contention that in the future, an integrated biomarker approach will be needed. The guidelines will have to separate the use of hsCRP measurements for purposes of primary prevention vs. secondary prevention. One problem, however, regarding specifically the “no other risk factor” subset, is that we have guidelines for dyslipidemia treatment, not for atherosclerosis as an inflammatory disease, guidelines for cholesterol problems, not for hsCRP problems. Do we need to widen the scope of our guidelines to accommodate inflammation-driven CVD risk?

Is JUPITER likely to result in an adjustment of risk stratification strategies?

Dr. Davignon: Yes, there is a need because at present: 1) we do not have a better substitute than hsCRP for recognizing patients at risk who have no elevation of LDL-C and few if any of the standard risk factors; 2) from the Reynolds risk evaluation score, evidence indicate that addition of hsCRP to the Framingham algorithm improves detection of individuals at risk; 3) equivalent risk is shared by those who have high LDL-C and normal hsCRP and those who have high hsCRP and normal LDL-C; 4) the bulk of the evidence indicates that hsCRP adds prognostic information beyond the Framingham risk score in all major cohorts evaluated, although some debates exist about this issue in a few studies. It will help bring to treatment people at intermediate risk who have high hsCRP and guideline-accepted LDL-C levels.

Dr. Fort: JUPITER confirms that knowing the patient’s hsCRP can be useful in specific clinical situations. Some individuals are already measuring hsCRP, but it has never been a required factor for risk assessment outside of a clinical trial. I will now be using it to evaluate risk in patients that would have been eligible for JUPITER. There is a risk it may be overused but, because this information is not useful in all patients, it should be used selectively.

Dr. McPherson: Again, hsCRP will be most useful in the borderline patient. We should not screen for hsCRP unless it is likely to change our management. In other words, we do not need hsCRP in very high-risk patients who need aggressive lipid lowering independent of hsCRP levels and we do not need hsCRP in very low-risk patients. In these patients, treatment is likely to be only marginally beneficial in the short term. However, in patients with risk factors but low LDL, an elevated hsCRP could provide an important rationale for statin treatment.

Dr. Verma: Conceptually it should, but we will have to wait and see what the guidelines recommend. JUPITER tells us that individuals who also had low risk (<10% by Framingham) had a similar benefit to those who had a >10% risk. A similar observation was made in those individuals with or without a family history. Therefore, conventional risk prediction algorithms will need to be re-calibrated to incorporate hsCRP within that paradigm. I believe that the Reynolds risk score for women and men is one superb attempt at this.

Dr. Genest: The evidence shows that hsCRP is going to be most helpful in the intermediate risk group. In the high-risk group patients, we know they need intensive lipid lowering so we do not need hsCRP. In low-risk groups, we do not have evidence that measuring hsCRP is beneficial, but in the intermediate-risk groups, hsCRP may help us make treatment decisions.

Dr. Anderson: I thought that the editorial that appeared in the same issue of the NEJM as the JUPITER results was thoughtful. Even though previous studies have suggested subjects with low hsCRP and normal cholesterol values have a low incidence of disease, one cannot exclude the possibility that statin therapy would not have reduced events in that cohort. The large reductions in LDL-C are surely important for the beneficial effects observed in the study; however, the specific role of hsCRP in the risk stratification is somewhat less clear. Moving forward, the information from the study clearly identifies those with elevated hsCRP as subjects who will gain benefit from aggressive LDL lowering.

Dr. Abramson: JUPITER did not demonstrate that hsCRP is a treatable risk factor: it is, at this point, only a risk marker. As a marker, hsCRP isolates a group of patients who will benefit from aggressive lipid lowering, but I am concerned that clinicians will think they should be screening hsCRP in all patients. This is not the case. In the older patients enrolled in JUPITER, an elevated hsCRP does isolate those who can reduce their CV risk through lipid lowering.

Can results of JUPITER be applied immediately to clinical practice?

Dr. Verma: Yes, provided hsCRP tests are readily accessible to physicians. The number needed to treat is 25, which is, in fact, lower than that for the use of statins to treat hyperlipidemia. This has impressive and long-reaching public health implications.

Dr. Davignon: Absolutely yes, but how to do it optimally and implement it is not obvious to me at present. I have been using hsCRP at our tertiary care Lipid Clinic for a few years as a supplemental tool of risk evaluation in all new patients referred to me, but have not yet reviewed the full value of this over time. Patients with no other CVD risk factors than hsCRP would not be seen at a lipid clinic unless they came with a positive family history or symptoms of vascular disease and seeking lipid evaluation, or by chance in a spouse in the context of a family survey.

Dr. Genest: The results of JUPITER are immediately relevant, but I think guidelines will be useful for determining how they should influence daily practice.

Dr. Fort: It is unusual for clinical practice to change on the basis of a single trial, but given the magnitude of this study, I think it will change practice for some even before the results are considered in the context of guidelines. From society’s perspective, I think the cost of screening for hsCRP and increasing the number of patients treated with statins is a concern, but routine hsCRP screening is going to be used more frequently to further maximize the benefits of statin therapy in the primary prevention population.

Dr. McPherson: I do think that physicians should consider using hsCRP in providing additional information when there is still doubt about whether to initiate treatment in patients assessed with conventional risk factors. When they become available, the next version of lipid treatment guidelines should further clarify the role of routine hsCRP measurements.

Dr. Anderson: These results are probably going to make many clinicians much more aggressive in terms of who they treat and how low they go with LDL. The target LDL in those enrolled in JUPITER would only be 3.5 mmol/L and we could see substantial benefits from going lower. As a cardiologist, I do not see many of the types of patients who were enrolled in JUPITER, so these results should be of greatest interest to primary care physicians. The benefit was robust enough that [the results] should influence clinical choices.

Dr. Abramson: Many physicians are likely to consider evaluating hsCRP as an additional factor to consider in patients at borderline risk for whom they are considering treatment. I think JUPITER also provides some reassurance regarding the safety of very low LDL-C levels. These results will certainly change the way we think about several clinical issues.

JUPITER included substantial numbers of women and minorities. Is it reassuring that the findings are applicable across all patient populations in the age ranges studied?

Dr. Abramson: The results of JUPITER justify my approach to treating women. The relatively small proportions of women included in previous statin trials have always left some in doubt. However, these data close the door on any concerns. I would expect women to be more similar rather than different relative to men in the pathophysiology of vascular disease. These results show conclusively that intensive lipid lowering is just as effective in women as in men.

Dr. Davignon: A resounding yes. This is where JUPITER results also come at an opportune time when value of statin treatment in women and the elderly is contested. It does not help with the criticism raised that statins might not be helpful in women before the menopause, though. For minorities, it has not been a major issue for me except for Asians, who might be more sensitive to statin side effects and respond differently to statins. [JUPITER results] do add strength to results acquired in the past for other minorities.

Dr. Fort: We have always had too little data regarding the use of statins in women and in minorities. I had expected the benefits to be similar in these groups, but now we have good evidence that it is. The study is important from that regard. Even at very low levels of LDL-C, there was comparable benefit in every group evaluated.

Dr. McPherson: JUPITER included close to 7000 women. I think this is extraordinarily important. Although there was no reason to suspect that women would benefit any less than men from statins, the lower number of women in previous statin trials always left some room for doubt. In JUPITER, we see that if anything, women may have benefited a little bit more. I think this finally puts to rest any doubt about the benefits of statins in women, minorities, or others who were poorly represented in previous studies.

Dr. Anderson: This was a powerful and welcome design feature. Although the benefits of lipid lowering in women and other subgroups have been reasonably robust in previous trials, these groups have generally been entered in small numbers. This provides the data for concluding that the benefits of statins are generalizable to all individuals.

Dr. Genest: There have been assertions made previously that women might not benefit as much as men. I do not think there has been much support for these assertions, but JUPITER randomized very large numbers to provide very convincing results. The same is true for non-Caucasian patients.

Dr. Verma: JUPITER is a landmark [trial] for this reason as well. It is the largest statin study in primary prevention in women and ethnic minorities, and as a result has important and generalizable implications.

The median LDL on treatment in JUPITER was 1.4 mmol/L. Do these results confirm that LDL levels this low are safe?

Dr. Genest: Previous data from randomized trials have provided strong evidence that very low LDL levels are safe. I did not need JUPITER for reassurance, but the very large numbers of patients enrolled in JUPITER should be reassuring to anyone who still had doubts.

Dr. Abramson: I have not been concerned about very low LDL-C levels. Other trials have looked for safety issues in patients achieving very low LDL-C levels and have not identified any apparent risks. [JUPITER] is another piece of evidence. It is important to recognize that the median follow-up in this study was less than two years, but the observed safety is consistent with that of previous studies.

Dr. McPherson: The subset analyses from TNT and PROVE-IT among other statin trials have also suggested that very low LDL levels are safe. The JUPITER data, in which about 25% of patients had a LDL on treatment of less than 1 mmol/L, provide even more reassurance that we do not need to be concerned when very low LDL levels are reached.

Dr. Fort: I hope this encourages aggressive LDL-C reductions even beyond current targets, but it is important to recognize that the evidence for safety of very low LDL-C, as seen in JUPITER, is only for the short or medium term. Median follow-up in JUPITER was only 1.9 years. Longer follow-up is needed for a more definitive answer.

Dr. Davignon: This is likely the case if we limit the answer to the period of observation, and if we look at the experience acquired from large studies that were carried out for a longer period of time and achieved similar very low levels (PROVE-IT) but not beyond. There is always the caveat that harm might occur in a particular subset with adverse ecological factors or with an adverse genetic predisposition. For example, it has been recently shown by Regieli et al. (Eur Heart J 2008;29:2792-9) that there is a 10-year increase in CVD mortality in statin-treated patients who are carriers of the TaqIB-B2 allele (HR 1.59 per copy) as compared to non-carriers. This is paradoxical, given that the B2 allele is associated with lower CETP and higher HDL-C levels. Another point marked for pharmacogenetics in predicting long-term drug effect.

Dr. Anderson: For a number of reasons, I have never been concerned about very low LDL-C levels. In populations with very little fat intake, LDL-C levels <1.0 mmol/L are common with no apparent ill health effects. Very low levels of LDL-C have also appeared safe in the subpopulations of patients who had LDL-C <1.0 mmol/L in previous statin trials. The JUPITER study showed a very clean risk profile for aggressive lipid lowering in a large number of patients. This cannot be considered the final answer because the follow-up was short, but there were no signals of any kind to make us nervous.

Is hs-CRP a reliable and consistent measure or are there circadian rhythms, dietary influences or other issues that must be considered? In other words, how easy is hs-CRP as a clinical tool?

Dr. Verma: hsCRP is very stable, easily assayed, and not altered by circardian rhythms, age or dietary influences. In patients who have active infection, hepatitis, arthritis, are taking HRT, steroids or other immunosuppressants, hsCRP may function as an acute phase reactant and in these populations, hsCRP testing may not be sensitive for CV events per se. It is important to remember that hsCRP has the similar stablility characteristics as blood pressure or even cholesterol.

Dr. McPherson: hsCRP levels do rise in a variety of chronic inflammatory diseases, such as scleroderma, rheumatoid arthritis and lupus, as well as in acute infections. In a patient without a chronic inflammatory disease and no signs or symptoms of an acute illness, two measurements at least two weeks apart are recommended, with the lower value being used for decision-making. The assays are well standardized and in otherwise healthy subjects, there is little variation over several months.

Dr. Davignon: I have alluded to this in the question about guidelines that must account for the introduction of this parameter with the perspective of the results of JUPITER: the need to establish to what extent hsCRP measurements are available and standardized across the country, the need for consensus regarding cut-off levels in the Canadian population, and the need to inform on the multiple sources of variation to improve the decision to measure (in whom and when) and the interpretation of results. hsCRP levels are highly modulated (age, exercise, genes, etc.) and also modulate many parameters (impact of other risk determinants, response to drugs, etc.).

Dr. Anderson: In studies that have looked at serial hsCRP measurements, about 25% will change risk category when re-evaluated. This is due to conditions that influence hsCRP, but the assays are reasonably standardized at this point, and hsCRP can be measured reliably if controlling for confounding factors such as infection or inflammatory diseases. This is a fairly simple tool.

Dr. Genest: The available assays appear to be relatively accurate. The elevations of hsCRP in patients with an acute infection or inflammatory disease are generally great enough that they will not be confused with the much more subtle increases in hsCRP associated with CV risk, so these can be performed reliably in most patients.

Dr. Abramson: I have not been measuring hsCRP routinely in my practice, but my understanding is that currently available assays are reliable. Certain diseases, including acute infections, can increase hsCRP, so it is appropriate to be cautious when interpreting results. I think it may help some physicians to have some specific guidelines on how best to measure hsCRP in routine practice.

Dr. Fort: The current hsCRP assays are reliable, but I think guidelines are important for us to know how to use it and when to use it. In addition to precautions to avoid false positive elevations, hsCRP will need to be employed selectively in patients for whom the result will change management. However, I think with appropriate guidelines, hsCRP is a tool that can be used by both general practitioners and specialists. It is a relatively simple and cheap test.