|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2008.20.0006 on March 30 2009 © 2009 American Society of Clinical Oncology.
Reply to H. Zetterberg et alDivision of Hematology/Oncology, Department of Medicine, University of California, San Francisco, CA
PPD Biomarker Discovery Sciences, Menlo Park, CA We recently used high-resolution proteomic technology to identify CSF biomarkers for focal brain lesions with an emphasis on lymphoma. We identified numerous differentially expressed CSF proteins associated with malignancy in individual patients: Approximately 80 CSF proteins were identified and found to be present at significantly different concentrations, both higher and lower, in training and test studies that were highly concordant. In addition, we used complementary methodology, including Western blot, immunohistochemistry, and enzyme-linked immunosorbent assay, for cross-validation of potential biomarkers. We focused on antithrombin III (ATIII) because of independent transcriptional profiling data that demonstrated high relative levels of RNA transcripts encoding this serine protease inhibitor within the tumors of CNS lymphoma patients, particularly those who had exhibited short survival. This study demonstrated, to our knowledge for the first time, that proteomic technology facilitates the discovery of individual biomarkers such as CSF ATIII that exhibit greater sensitivity than does CSF cytology in the identification of cancer.1 Zetterberg et al2 now raise the question of whether CSF ATIII can be considered a CNS lymphoma biomarker, given that a significant proportion of ATIII in the CSF seems to enter the CSF through disruption of the blood-brain barrier. First, it should be noted that the conclusions of Zetterberg et al2 are based on the analysis of CSF specimens from a cohort of 38 patients without CNS lymphoma or other brain tumors. Inclusion of such patients would seem essential to support their thesis. Given our demonstration that ATIII is expressed by CNS lymphoma, we believe that this fact especially weighs against their argument. Had they included CSF from brain tumor patients in their analysis, they would have realized that ATIII levels reflect two biomarker changes as a result of cancer, a direct tumoral biomarker as well as one reporting on the effect of cancer on the blood-brain barrier. This position is directly supported by additional data, generated by measuring the concentration of CSF ATIII and CSF albumin (Human Albumin ELISA Quantitation Kit, Bethyl Laboratories, Montgomery, TX) from specimens obtained from CNS lymphoma and control patients. This analysis demonstrates that CSF ATIII is a significantly more accurate biomarker than CSF albumin in discriminating patients with CNS lymphoma. We compared the differential concentration of CSF ATIII with CSF albumin among 53 CSF specimens from 28 patients in a control group, many of whom were being evaluated for possible CNS lymphoma, and 25 patients with CNS lymphoma. The mean CSF ATIII concentration in the patients with lymphoma was 3.1 µg/mL; the mean CSF ATIII concentration among patients in the control group was 0.53 µg/mL. As expected, the average concentration of albumin in the CSF was also higher in the patients with lymphoma relative to patients in the control group (560 µg/mL v 140 µg/mL). Receiver operating characteristic analysis3 demonstrated that the diagnostic accuracy of CSF ATIII (area under the curve [AUC] = 0.92) is superior to the diagnostic accuracy of either CSF albumin (AUC = 0.74) or CSF total protein (AUC = 0.75) in distinguishing patients with lymphoma compared with patients in the control group (P < .002; Fig 1).
We also cite data, which were provided to the referees during the review process of our report by Journal of Clinical Oncology, that demonstrate that the CSF concentration of ATIII is elevated when normalized to total protein in CNS lymphoma patients compared with patients in the control group. These data support our evidence for the intratumoral expression of ATIII, which was demonstrated by transcriptional profiling analysis. Paired CSF and total protein concentrations, available in more than 80% of the 101 patients analyzed, demonstrated that CSF ATIII is significantly elevated in patients with brain tumors compared with patients in the control group, even when normalized to total CSF protein: Mean normalized CSF ATIII concentration in control group patients was 0.0014; mean normalized CSF ATIII concentration CNS in lymphoma patients was 0.0021; and mean normalized CSF ATIII concentration in glioma/brain metastasis patients was 0.0023. The P value for the difference between mean normalized CSF ATIII concentration in patients in the control group versus patients with CNS lymphoma is less than .015 (t-test). The P value for the difference between mean normalized patients in the control group versus patients with glioma/brain metastasis is less than .018 (t-test). It is also important to point out that the position adopted by Zetterberg et al2 is essentially that the biomarker must originate in the cancer tissue, and not as a pathologic consequence of the cancer. This is in disagreement with the formal definition of biomarker, according to the dictionary of cancer terms of the National Cancer Institute: "Biomarker: a biologic molecule found in blood, other body fluids, or tissues that is a sign of a normal or abnormal process, or of a condition or disease. A biomarker may be used to see how well the body responds to a treatment for a disease or condition. Also called molecular marker and signature molecule." We believe that CSF ATIII has diagnostic utility both as a consequence of its synthesis by the tumor and as a consequence of disruption of the blood-brain barrier, one of the integral signs of the disease. We have used two-dimensional liquid chromatography/mass spectrometry to identify a multitude of potential CNS lymphoma protein biomarkers that are differentially expressed in the CSF; among these, ATIII has been validated by multiple technologies and fulfills criteria as a novel biomarker of CNS lymphoma with significantly better diagnostic utility compared with the measurement of CSF albumin, total CSF protein, or CSF cytology. AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: Sushmita Roy, PPD (C); Chris Becker, PPD (C); Howard Schulman, PPD (C) Consultant or Advisory Role: None Stock Ownership: None Honoraria: None Research Funding: None Expert Testimony: None Other Remuneration: None ACKNOWLEDGMENT We thank Ken Fujimura and Valerie Wong for their technical assistance. REFERENCES
1. Roy S, Josephson SA, Fridlyand J, et al: Protein biomarker identification in the CSF of patients with CNS lymphoma. J Clin Oncol 26:96–105, 2008. 2. Zetterberg H, Andreasson U, Blennow K: CSF antithrombin III and disruption of the blood-brain barrier. J Clin Oncol doi:10.1200/JCO.2008.19.8598. 3. Zweig MH, Campbell G: Receiver-operating characteristic (ROC) plots: A fundamental evaluation tool in clinical medicine. Clin Chem 39:561–577, 1993.
Related Article
Related Correspondence
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2009 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
