| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Department of Pathology Johns Hopkins Medical Institutions Baltimore, MD
aAddress correspondence to this author at: Department of Pathology, Johns Hopkins Medical Institutions, 419 North Caroline St., Baltimore, MD 21231. Fax 410-502-7882; e-mail jli{at}jhmi.edu.
To the Editor:
We thank Dr. Diamandis for his comments and wish to comment on some of the issues that he raised.
First, we want to clarify that it was not the intention of the original discovery paper (1) or of the recent validation paper (2) to determine the suitability of "this [SELDI] technology for cancer diagnostics". We used surface-enhanced laser desorption/ionization (SELDI) technology for biomarker discovery and validation.
We do not disagree with Dr. Diamandis’ statement and prediction that candidate biomarkers discovered in serum or plasma by SELDI (and by many other high-throughput proteomics profiling technologies) tend to be high-abundance proteins. We believe, however, that this situation reflects not only the low sensitivity of the current mass spectrometers and other detection technologies, but also the lack of sample preparation (depletion, fractionation) technologies that are applicable for high-throughput analysis with high reproducibility (low CV in both mass accuracy and relative protein expression measurement).
The letter by Dr. Diamandis did not provide any actual data to discount the findings in our 2 papers; the questions he raised centered on the more general issues of whether such abundant proteins can truly be disease-associated biomarkers and whether they can be clinically relevant.
Proteolytic processing of peptides in circulation is a well-known phenomenon, and many of the protein fragments are from abundant proteins that are less likely to be directly secreted from the actual tumor sites. It is also possible, however, that cancer cells and/or the host immune system may produce proteases and other enzymes that are tumor specific and may produce specific fragmentation and cleavage of common proteins, including abundant proteins. An increasing body of evidence, from our own group (3)(4) and from others (5)(6), supports such possibilities. The fact that many known serum/plasma biomarkers are abundant proteins or their derivatives could be caused by limitations of the current technology. Alternatively, such biomarkers could be the "amplified" signal of a disease process that otherwise might not be easily detected at an early stage. A tumor biomarker is evaluated by its relevance to the disease; it is not a prerequisite for it to be a direct product of the tumor.
Cancers are heterogeneous diseases with multiple subphenotypes, each following a distinct pathway. With possibly a few exceptions, it is unlikely that a single marker will provide acceptable diagnostic sensitivity and specificity for any of these complex diseases. It is our firm belief that multiple biomarkers, each with clinically verifiable relevance to the disease, combined through statistically sound approaches will offer better diagnostic performance than any individual marker. It is possible that some biomarkers, such as BC2 and BC3, may not be super performers on their own, but provide value in combination with others.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
