Individualized Medicine by Biomarkers
Individualized Medicine by Biomarkers
Abstract & Commentary
By Michael H. Crawford, MD
Source: Melander O, et al. Novel and conventional biomarkers for prediction of incident cardiovascular events in the community. JAMA. 2009;302:49-57.
After the JUPITER trial showed that individuals with LDL cholesterols < 130 mg/dL, but high sensitivity CRP > 2.0 benefited from statin therapy, there has been considerable interest in using biomarkers for individualizing treatment. Those at low risk for cardiovascular events (CVE) by traditional risk factors are a significant portion of those that actually have an event because of their large numbers. Thus, this study of largely low- to intermediate-risk people without known coronary artery disease, from Sweden, is of interest. Melander et al studied CRP, BNP, cystatin C, Lp-PLA2, MR-proADM (multiregional pro adrenomedullin), and MR-proANP in about 5,000 subjects, 60% of whom were women, with a mean age of 58 years. They compared the predictive value of these biomarkers, alone or in combination, against traditional clinical risk factors such as the Framingham score over a median follow-up of 13 years. The primary endpoints were myocardial infarction or ischemic events (CE) and CE plus stroke (CVE). Secondary endpoints were total mortality and CVE plus heart failure. Events were classified by ICD-9 codes. Advanced predictive statistics were used.
Results: There were a total of 418 CVE (8%) and 230 CE (4%) over 10 years. The predictive analysis of single biomarkers for CVE showed that five of the six measured were significant (Lp-PLA2 dropped out), but the hazard ratios (HR) were modest (1.1-1.2). Only three predicted CE (cystatin C, MR-proADM and BNP), with slightly higher HRs (1.2-1.3). Also, biomarkers individually added minimally to the HRs of the Framingham risk scores (all NS for CVE). Only MR-proADM was significant for CE (p = .02). When multiple biomarkers were considered, CRP was predictive of CVE, MR-proADM was predictive of CE, and BNP was predictive of both types of events, but the incremental gain was small and only 8% of patients were reclassified, most to a lower-risk range. Melander et al concluded that certain biomarkers are predictive of cardiovascular events, but the incremental gain over traditional risk factors is minimal.
Commentary
Previous studies of the risk prediction abilities of biomarkers have shown good results in higher-risk individuals, but less value in low- to intermediate-risk individuals. The latter are where the need for more accurate risk prediction is greatest, since these lower-risk patients are most likely to be impacted by the results. Also, there is little data on the use of multiple biomarkers, yet we know several are predictive individually. This study targeted intermediate-risk individuals and evaluated six biomarkers. Perhaps the most interesting finding is that Lp-PLA2 was not predictive in these patients. This marker is supposed to be more specific than CRP and identify the presence of vulnerable plaques. Either it does not do that, or there were too few plaque rupture events to determine its predictive power.
Basically, this was a negative study from a clinicians' point of view. Although five of the six biomarkers were predictive of events, the hazard ratios were modest (1.1-1.4), and few patients were reclassified compared to their traditional risk factor classification. Also, most of those reclassified were to a lower-risk class, which does not help the clinician justify more aggressive therapy. In fact, by ATP III criteria, only 1% of these patients would warrant a therapy change based upon the biomarkers result.
Why was this study different from previous studies? The most striking difference was that 60% of the participants were women and, not surprisingly, the average HDL cholesterol was 54 mg/dL. LDL-cholesterol and systolic blood pressure mean values (162 mg/dL and 141 mm/Hg) were solidly intermediate risk, but the high proportion of women may have skewed the study toward lower risks. However, the 10-year total mortality of about 8% is low to intermediate risk. Another issue is that it is difficult to show incremental gains, especially when traditional risk factors already do a fairly good job of classifying individuals. Finally, we may not have the ideal biomarkers yet. Of the six, BNP seemed the most valuable overall, yet clinicians know about the problems with interpreting a positive result. The same can be said for a high CRP value; it is most often a false positive. These most widely used biomarkers are clearly of more clinical value if they are negative or normal. Thus, it is not surprising that the majority of classification changes in this study were to lower risk levels based upon a negative biomarker. At this point, the role of biomarkers in personalized cardiovascular medicine is unclear.
After the JUPITER trial showed that individuals with LDL cholesterols < 130 mg/dL, but high sensitivity CRP > 2.0 benefited from statin therapy, there has been considerable interest in using biomarkers for individualizing treatment.Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.