|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2008.21.6127 on February 23 2009 © 2009 American Society of Clinical Oncology.
In ReplyMemorial Sloan-Kettering Cancer Center, New York, NY We would like to think Nanda et al for their interest in our article. We are in strong agreement with their view that evaluation of markers must go beyond measures of accuracy and take into account clinical relevance. Indeed, one of us has written extensively on this point1–3 and has developed a novel statistical methodology for assessing the clinical value of markers.4,5 We have applied this methodology to evaluate markers that we have been developing.6 In particular, we have shown that the methodology is a decision-analytically weighted variant of reclassification,2 which is what Nanda et al seem to endorse in their letter. Nanda et al cite three articles which show that prostate-specific antigen (PSA) velocity can help predict outcomes in men categorized as low risk by absolute PSA level. For example, in one of the cited articles, the authors examine 125 men defined as being at low risk and show that there were more recurrences in men with a high PSA velocity than in men with a low PSA velocity.7 This suggests that PSA velocity might be useful for reclassification. The problem is that the classification scheme used is rather a crude one: men are classed at low risk if they have a PSA less than 10 ng/mL; most urologists would say that a man with a PSA of 9.9 ng/mL was at higher risk than a man with a PSA of 3.0 ng/mL. As a simple demonstration of why this is a problem, we analyzed some of our own data using a similar method to the cited article, except that instead of PSA velocity being the criterion for high risk, we used a PSA of 6 ng/mL or above. Five-year recurrence-free probability was 96% for men with low PSA but 90% for men with high PSA (P = .007). In other words, absolute PSA level was able to identify men at high risk of recurrence despite being categorized as low risk by absolute PSA level. The paradoxical nature of this sentence underlines that the analysis is a problematic one. None of the articles we reviewed included any decision-analytic method that addressed the question of whether clinical decision making is improved by incorporation of PSA dynamics. Accordingly, we stand by our original conclusions. Finally, Nanda et al contrast detection and prognosis, comparing "The potential utility of PSA dynamic constructs in men with a diagnosis of [prostate cancer] as opposed to the general population where benign causes can spuriously elevate PSA (ie, prostatitis, instrumentation, sexual activity, bicycle or horseback riding)." Our view is that the same confounders also affect men recently diagnosed with prostate cancer. Nonetheless, with respect to their request that detection and prognosis "be considered separately," this is obviously easy to do: if there is little evidence for PSA dynamics when both clinical situations are combined, there is even less when each is considered separately. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST The author(s) indicated no potential conflicts of interest. REFERENCES 1. Vickers AJ, Jang K, Sargent D, et al: Systematic review of statistical methods used in molecular marker studies in cancer. Cancer 112:1862–1868, 2008.[CrossRef][Medline] 2. Vickers AJ, Elkin EB, Steyerberg E: Net reclassification improvement and decision theory. Stat Med 28:525–526, 2009.[CrossRef][Medline] 3. Vickers AJ: Decision analysis for the evaluation of diagnostic tests, prediction models, and molecular markers. Am Stat 62:314–320, 2008.[CrossRef][Medline] 4. Vickers AJ, Elkin EB: Decision curve analysis: A novel method for evaluating prediction models. Med Decis Making 26:565–574, 2006. 5. Vickers AJ, Cronin AM, Elkin EB, et al: Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers. BMC Med Inform Decis Mak 8:53; 2008.[CrossRef][Medline] 6. Vickers AJ, Cronin AM, Aus G, et al: A panel of kallikrein markers can reduce unnecessary biopsy for prostate cancer: Data from the European Randomized Study of Prostate Cancer Screening in Goteborg, Sweden. BMC Med 6:19; 2008.[CrossRef][Medline] 7. D'Amico AV, Renshaw AA, Sussman B, et al: Pretreatment PSA velocity and risk of death from prostate cancer following external beam radiation therapy. JAMA 294:440–447, 2005.
Related Articles
Related Correspondence
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2009 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
