| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Imaging, Diagnosis, Prognosis |
Authors' Affiliations: Departments of 1 Medical Oncology and 2 General Surgery I and II; 3 Chemical-Clinical Analysis Laboratory; 4 Blood Transfusion Unit, Morgagni-Pierantoni Hospital; 5 Istituto Oncologico Romagnolo, Forlì, Italy; and 6 Istituto Scientifico Romagnolo per lo Studio e la Cura dei Tumori, Meldola, Italy
Requests for reprints: Emanuela Flamini, Department of Medical Oncology, Morgagni-Pierantoni Hospital, via Forlanini 34, 47100 Forlì, Italy. Phone: 39-0543-731840; Fax: 39-0543-731736; E-mail: e.flamini{at}ausl.fo.it.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Experimental Design: The study was carried out on 75 healthy donors and 75 colorectal cancer patients. Free DNA was determined in serum with quantitative PCR analysis. The diagnostic accuracy of each assay was calculated using receiver operating characteristic (ROC) curves. The diagnostic relevance of the two-marker combination was analyzed by the logistic regression model.
Results: Median free DNA concentration was 
5-fold higher in patients than in healthy donors (P < 0.001). The area under the ROC curve was 0.86, and when 12.5 ng/mL was used as cutoff, 81.3% sensitivity and 73.3% specificity were observed for the overall series. As CEA and free DNA provided independent diagnostic information, they were also considered in combination. ROC curve analysis of the combined CEA and free DNA algorithms showed a higher diagnostic capacity (area under the ROC curve, 0.92) than that of markers considered singly, with 84% sensitivity and 88% specificity.
Conclusions: Free circulating DNA, especially when used in combination with CEA, represents a potentially useful tool for the diagnosis of early-stage colorectal cancer.
The most widely used screening technique for colorectal cancer is the fecal occult blood test. However, this simple, inexpensive, and noninvasive test is heavily prone to produce not only false-positive results but also false-negative results because colorectal tumors bleed intermittently.
On the other hand, colonoscopy, which has very high diagnostic accuracy in terms of both sensitivity and specificity, is characterized by a moderate compliance because it is invasive and not without potentially adverse events. Its use is limited to second-level diagnostic tests within screening programs (3–5).
Numerous serum markers, such as carcinoembryonic antigen (CEA), carbohydrate antigen 19-9, and lipid-associated sialic acid, have been investigated in colorectal cancer, but their low sensitivity has induced the American Society of Clinical Oncology to state that none can be recommended for screening and diagnosis and that their use should be limited to postsurgery surveillance (2, 4, 6–8).
The search, therefore, continues for markers that match the diagnostic accuracy of fecal occult blood test but are less invasive than colonoscopy for the early detection of colon cancer.
Free circulating DNA has produced interesting results for lung and breast cancer (9–12). The few studies of this marker done for colon cancer have shown higher levels of circulating DNA in patients than in healthy individuals (13–15), but its diagnostic relevance has not been adequately investigated. The biological characterization of circulating DNA in the blood of patients has shown that an important component derives from tumor cells (16, 17). Although the mechanism of DNA release into the blood is unknown, cell apoptosis or necrosis, as well as the active release of DNA, have been hypothesized (18).
The aim of the present study was to define the relevance of a relatively simple and inexpensive test based on circulating DNA alone and in combination with the more conventional CEA analysis (6–8) for the early detection of colorectal cancer.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The median age of healthy individuals was 57 years (range, 29-79 years) and that of patients was 71 years (range, 38-89 years). The control group comprised 27 females and 48 males, whereas 32 females and 43 males made up the patient group.
Twenty-two cancers were located in the right colon, 38 in left colon, 3 in the transverse colon, and 12 in the rectal tract. Tumors were classified according to Dukes' staging system: 18 were stage A, 19 were stage B, 26 were stage C, and 12 were stage D.
The study protocol was reviewed and approved by the local ethics committee in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. All subjects gave their informed written consent to take part.
Sample collection
A 5-mL venous blood sample from patients presurgery and from controls was collected in tubes without anticoagulant and centrifuged at 2,500 rpm for 15 minutes at room temperature. Serum was then stored at –80°C until processing.
Marker determination
Carcinoembryonic antigen. The CEA assay was done on serum in the Clinical Pathology Laboratory of Morgagni-Pierantoni Hospital using chemiluminescent microparticle immunoassay technology (Abbott Laboratories, Abbott Park, IL). The Architect CEA assay consists of a two-step quantitative immunoassay. The limit of normality was considered as 5 ng/mL.
Free DNA. Serum samples were processed within a maximum of 2 months to avoid biases caused by prolonged sample storage. DNA extracted from 1 mL serum was purified according to the protocol used by Sozzi et al. (9, 10) and stored at –20°C.
To quantify circulating free DNA, the glyceraldehyde-3-phosphate dehydrogenase housekeeping gene, used recently in a study on lung cancer (11), was amplified by a real-time quantitative PCR assay (19) using SYBR Green I Dye Chemistry (20, 21) and MyIQ Single-Color Real-time PCR Detection System (Bio-Rad, Hercules, CA). Forward and reverse primer sequences were 5'-ACCCAGAAGACTGTGGATGG-3' and 5'-TTCAGCTCAGGGATGACCTT-3', respectively. PCR mix was prepared in a total volume of 25 µL containing 1x SYBR Green Supermix (Bio-Rad), 0.4 µmol/L of each primer, and 5 µL DNA. PCR conditions were set up as follows: a first denaturation at 95°C for 3 minutes and 40 cycles for 15 seconds at 95°C and 30 seconds at 55°C for annealing and extension.
The absolute concentration of target DNA was calculated on a standard curve using concentrations ranging from 25 to 0.01 ng DNA extracted from the peripheral blood of a healthy donor. Each sample was run in triplicate and the intra-assay variability was assessed by computing the coefficient of variation among the three Ct values (defined as the fractional cycle number at which the emitted fluorescence exceeds a fixed threshold value above the baseline), which was always <1.5%. The interassay variability between two independent experiments was assessed by the coefficient of variation and was always <15%.
Statistical analysis
Using the best cutoff value, the population size was defined based on an expected sensitivity and specificity of 75%. For 75 healthy donors and 75 cancer patients, we predicted the 95% confidence intervals (95% CI) to be ±10% with respect to the single estimated value for sensitivity and specificity.
The median values of each marker in controls and cases were compared using the median test. The correlation between serum levels of the two markers was assessed with the Spearman rank test. In the absence of internationally available cutoff values for serum free DNA, the cutoff maximally discriminating between healthy donors and patients was identified using receiver operating characteristic (ROC) curve analysis (22). The 95% CI values were calculated for sensitivity and specificity values. The independent diagnostic relevance of the two markers considered as continuous variables (subjected to natural logarithmic transformation) was analyzed by the logistic regression model (23).
The linear predictor or logit resulting from the model was used as a new diagnostic test on which the ROC curve was calculated. Statistical analysis was done using SPSS software.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
5-fold lower in healthy individuals (7.7 ng/mL; range, 0.2-46.6 ng/mL) than in patients (35.8 ng/mL; range, 0.6-153 ng/mL; P < 0.001). Median concentrations of the two markers were independent of age and gender in both donors and patients and did not differ as a function of clinical or pathologic features, with the exception of Dukes' stage (data not shown). In fact, median CEA levels were similar in Dukes' stage A to C tumors and increased in patients with stage D disease. The lack of an increase from stages B to C could be due to the higher percentage of grade 3 tumors in stage C (31%) compared with stage B (11%). Conversely, free DNA median levels were already high in patients with early-stage tumors (Tables 1
and 2
).
|
|
|
|
As CEA and free DNA were not significantly correlated in either donors (rs = 0.04; P = 0.77) or patients (rs = 0.16; P = 0.17), the two markers were analyzed simultaneously. When considered as dichotomous variables, positivity to at least one of the two markers resulted in specificity (70.7%; 95% CI, 59.7-79.5) similar to that of a single marker, whereas sensitivity increased (88%; 95% CI, 78.7-93.5). Furthermore, when both were positive, specificity reached 100%. When the markers were used as continuous variables, transformed into logarithmic values, and analyzed in a logistic regression model, each marker provided independent and therefore additive diagnostic information. From the model, a unit increase in log serum DNA or in log CEA was associated with an increase of 5-fold in cancer risk (Table 4
).The ROC curve analysis of the "combined" CEA and free DNA algorithms showed a higher diagnostic capacity than that of either marker considered singly (AUC, 0.92; 95% CI, 0.88-0.92; Fig. 1). The best cutpoint of this algorithm (0.18) was associated with 84% sensitivity and 88% specificity. Sensitivity of the two markers tested in combination was similar in patients with Dukes' stage A (83.3%; 95% CI, 60.4-93.9), B (89.5%; 95% CI, 68.3-96.8), and C (84.6%; 95% CI, 66.3-93.7) and reached 100% in patients with Dukes' stage D disease.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
CEA, the most widely used serum marker in colorectal cancer, offers a sensitivity of only 30% to 40% for early-stage tumors (6–8), also confirmed in our study. A large number of studies have reported higher concentrations of serum/plasma free DNA in patients with various types of cancer (9–12, 24). Seventy-eight percent sensitivity and 95% specificity for circulating free DNA levels have been reported for non–small cell lung cancer patients (10). In patients with breast cancer, higher free DNA levels have been observed compared with those of controls, independently of tumor size and stage (12). In the present study, median CEA levels were lower in patients with Dukes' stage A to C disease and increased in metastatic cancer, whereas high free DNA levels were observed starting from early-stage tumors.
As far as we know, ours is the first case-control study on colorectal cancer patients. Free DNA levels were 5-fold higher in the serum of patients than in that of controls and were not related to either age or gender. Moreover, circulating DNA showed 81.3% sensitivity, which increased to 88% when free DNA and CEA were considered in association. The combination of the two markers could therefore be a useful tool for the diagnosis of early-stage disease. Based on these findings, the free DNA test would seem to represent a rapid and relatively inexpensive alternative to the more invasive colonoscopy.
A comparative analysis of circulating free DNA and DNA amplification in feces, routinely carried out in our laboratory, is currently being done on the same case series to define the diagnostic accuracy of the single tests (25–27).
|
Acknowledgments |
|---|
|
Footnotes |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received 8/ 4/06; revised 9/ 6/06; accepted 9/14/06.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. F. Ransohoff, C. Martin, W. S. Wiggins, B. A. Hitt, T. O. Keku, J. A. Galanko, and R. S. Sandler Assessment of Serum Proteomics to Detect Large Colon Adenomas Cancer Epidemiol. Biomarkers Prev., August 1, 2008; 17(8): 2188 - 2193. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
