Background: The presence of circulating antibodies to the p53 tumor suppressor protein is a potential early detection colorectal cancer biomarker. However, studies of prediagnostic measures of p53 seropositivity in relation to colorectal cancer risk are limited.

Methods: We conducted a nested case–control study of serum p53 autoantibodies and risk of colorectal cancer within the Cancer Prevention Study-II Nutrition Cohort. Among cohort participants who were cancer free at the time of blood collection, 392 were subsequently diagnosed with colorectal cancer over 11 years of follow-up. Two controls were matched to each case on birth date, blood draw date, race, and sex. Autoantibodies to p53 were detected in 41 of the 392 cases (10.5%) and 49 of the 774 controls (6.3%).

Results: Participants who were seropositive for p53 antibodies before diagnosis were more likely to be subsequently diagnosed with colorectal cancer [RR = 1.77; 95% confidence interval (CI), 1.12–2.78]. This association was strongest within 3 years of diagnosis (RR = 2.26; 95% CI, 1.06–4.83). An association was also suggested when colorectal cancer was diagnosed 4 to <6 years after p53 measurement (RR = 1.84; 95% CI, 0.89–3.79), but not 6 or more years later (RR = 1.15; 95% CI, 0.44–2.99).

Conclusions: If these results are confirmed, serum p53 antibodies may be useful on a panel of early detection markers for colorectal cancer.

Impact: Individuals who were seropositive for p53 antibodies were twice as likely to develop colorectal cancer within the next 3 years compared with those who were seronegative. This marker is a good candidate for inclusion on an early detection marker panel for colorectal cancer. Cancer Epidemiol Biomarkers Prev; 27(2); 219–23. ©2017 AACR.

Colorectal cancer is a multifactorial disease with several established risk factors, including age, male sex, obesity, and lack of physical activity. Blood biomarkers have been investigated as potential early detection markers (1–4) and, if confirmed, may be useful in complementing current colorectal cancer screening protocols. Specifically, blood biomarkers could help identify people with undetected colorectal cancer.

The presence of circulating antibodies to the p53 tumor suppressor protein is a potential biomarker of colorectal cancer risk (3). Somatic mutations in the gene that encodes p53, TP53, are thought to be involved in about half of all human cancers (5), including colorectal cancer (6). The specific trigger for the appearance of p53 autoantibodies in serum is not known, but it is hypothesized to be related to p53 protein overexpression (4). Somatic TP53 mutations alone are insufficient to trigger anti-p53 autoantibody production, and only 20% to 40% of patients with somatic p53 mutations are seropositive for the antibody (5).

Regardless of the specific mechanism underlying the autoantibody response, an association between p53 antibody seroprevalence and cancer prevalence has been observed for many types of cancer. A summary article combining data from 80 serologic studies among cancer cases (n > 9,000) and 36 serologic analyses among noncases (n = 2,400) showed a strong association between p53 antibody seroprevalence and any cancer (16.9% among cases vs. 1.45% among controls, P < 10−4) including colorectal cancer (24.7%, P < 10−4; 5). A major limitation of these studies, however, is that p53 antibodies were measured after colorectal cancer diagnosis, leaving it unknown whether these antibodies would have been detectable before clinical diagnosis of the cancer. To our knowledge, only one other study examined the association between prediagnostic p53 autoantibody status and subsequent colorectal cancer risk; a prospective study of 97 colorectal cancer cases nested within the United Kingdom Collaborative Trial of Ovarian Cancer Screening suggested that elevated levels of p53 antibodies can be detected up to 4 years before clinical diagnosis of colorectal cancer among women (3).

To follow-up on the findings from the ovarian cancer screening study (3), we analyzed archived blood samples from women and men enrolled in the Cancer Prevention Study-II (CPS-II) Nutrition Cohort, a large prospective study of both men and women in the United States. Prediagnostic serum p53 antibodies were measured, and the association between this biomarker and risk of colorectal cancer was evaluated.

Study population

The CPS-II Nutrition Cohort is a prospective study of 184,194 men and women established in 1992 in the United States (7). Follow-up questionnaires have been sent to cohort members in 1997, and every 2 years thereafter, to ascertain newly diagnosed cancers and collect updated exposure information. Among participants who were mailed follow-up questionnaires, response rates were at least 87%. Between 1998 and 2001, a subset of CPS-II Nutrition Cohort participants was invited to provide a blood sample at a medical facility in their community. Nonfasting blood samples were obtained from 32,704 participants without a personal history of cancer at the time of blood collection; these participants also completed a brief questionnaire and provided informed consent. Blood samples were collected in two 15-mL EDTA tubes and a 13-mL serum separator tube. The tubes were kept chilled while shipped overnight to a central repository in Rockville, Maryland, where plasma and sera were removed after centrifugation, and then aliquoted and placed in liquid nitrogen freezers for long-term storage. The recruitment, characteristics, and follow-up of the CPS-II Nutrition Cohort are described in greater detail elsewhere (7). All aspects of the CPS-II cohort were approved by the Emory University Institutional Review Board.

Case ascertainment

Cases in this analysis (n = 369) were defined as individuals with no previous history of cancer who were diagnosed with colorectal cancer between the date of blood collection and June 30, 2009 (end of follow-up for this analysis). Colorectal cancer cases were identified by self-report on biennial questionnaires and subsequently verified through medical records (n = 277) or registry linkage (n = 92). An additional 23 cases were identified through National Death Index records.

Control selection

For each of the 392 eligible colorectal cancer cases, we randomly selected two matched controls from among participants who provided a blood sample, were alive, and had no personal history of cancer on the diagnosis date of the case. Each control was individually matched to their case on birth date (±6 months), blood draw date (±6 months), sex (male/female), and race (white, black, other/unknown). Six cases could not be matched to their controls on race but were matched on the other criteria.

Antibody measurements

Serologic analyses to measure antibodies to the cellular protein p53 were conducted at the German Cancer Research Center in Heidelberg, Germany. Frozen serum samples were shipped on dry ice. Seroreactivity against p53 antibodies was measured by fluorescent bead–based Multiplex Serology and expressed as median fluorescence intensity (MFI). Using this multiplex methodology, p53 antibodies were measured along with a panel of viral and bacterial antibodies. This method has been described in detail elsewhere (8, 9). Briefly, Homo sapiens p53 (NCBI Acc No: NM_000546) was cloned as full-length gene into expression plasmid pGEX4T3tag between N-terminal glutathione S-transferase (GST) and a C-terminal epitope (tag) derived from the large T-antigen of simian virus 40 (10). Fusion proteins were expressed in E. coli BL21 Rosetta (Novagen-Merck). Glutathione cross-linked to casein was covalently coupled to fluorescence-labeled polystyrene beads (SeroMap; Luminex), and GST-p53 fusion proteins were affinity purified on the beads directly. Bead types of different colors, each carrying a different antigen, were mixed with human serum and analyzed at 1:1,000 dilution. Human antibodies bound to the beads were stained with biotinylated goat anti-human IgA, IgM, IgG, and reporter conjugate streptavidin-R-PE. Beads were examined in a Luminex 100 analyzer that identified bead color and quantified the antibody bound to the p53 antigen via the median R-PE fluorescence intensity of at least 100 beads of the same internal color. MFI values of 200 or higher were classified as seropositive for p53 antibodies, and all lower values were seronegative.

A total of 49 replicate samples from 19 additional CPS-II participants were used as quality controls and assayed at the same time as the case and control samples. The overall coefficient of variation for these replicate samples was 16% and the corresponding intraclass correlation coefficient was 72%.

Statistical analysis

Rate ratios (RR) and 95% confidence intervals (CI) were estimated using conditional logistic regression. Models were conditioned on matched triplicate, thereby adjusting for the matching factors: sex, draw date, birth year, and race. Seropositivity as well as antibody levels were evaluated. For dose–response analyses, cutoff points for antibody levels were calculated from MFI tertiles among seropositive control participants.

Among the 1,166 serum samples tested, 90 were seropositive for p53 autoantibodies. Characteristics of the study population by p53 antibody serostatus are shown in Table 1. Those who tested positive for p53 autoantibodies were slightly more likely to be female, less educated, a current smoker, and to have a lower BMI, but none of these differences were statistically significant. Median age at blood draw was 72 years for the participants who were seropositive for p53 antibodies and 71 years for those who were seronegative (range for both, 56–87 years). Median age at diagnosis for the cases was 76 years (range, 60–91). Forty-one of the 392 colorectal cancer cases were seropositive for p53 autoantibodies (10.5%) compared with 49 of the 774 controls (6.3%).

Table 1.

Characteristics of the study population at the time of blood collection by prediagnostic p53 antibody serostatus

p53 antibodies
Seron = 1,076Sero+n = 90χ2P value
Median (range) age at blood draw (years) 71 (56–87) 72 (56–87)  
Median (range) years before diagnosis 4.1 (<1–10.8) years 3.4 (<1–9.8 years)  
Sex 
 Female 553 (51.4%) 51 (56.7%) 0.34 
 Male 523 (48.6%) 39 (43.3%)  
Race 
 White 1,057 (98.2%) 88 (97.8%) 0.93 
 Black 8 (0.7%) 1 (1.1%)  
 Other/unknown 11 (1%) 1 (1.1%)  
Education 
 Less than high school 36 (3.3%) 3 (3.3%) 0.57 
 High school/missing education 230 (21.4%) 25 (27.8%)  
 Some college/vocational school 340 (31.6%) 26 (28.9%)  
 College/graduate school 470 (43.7%) 36 (40.0%)  
BMI 
 <25.0 454 (42.2%) 45 (50.0%) 0.10 
 25.0–<30.0 431 (40.1%) 27 (30.0%)  
  ≥30.0 172 (16.0%) 14 (15.6%)  
 Unknown 19 (1.8%) 4 (4.4%)  
Smoking status 
 Never 485 (45.1%) 37 (41.1%) 0.60 
 Current 28 (2.6%) 4 (4.4%)  
 Former 549 (51.0%) 47 (52.2%)  
 Unknown 14 (1.3%) 2 (2.2%)  
Alcohol intake 
 Nondrinker 261 (24.3%) 24 (26.7%) 0.76 
 ≤1 drink/day 563 (52.3%) 44 (48.9%)  
 >1 drink/day 244 (22.7%) 22 (24.4%)  
 Unknown 8 (0.7%) 0 (0%)  
Physical activity 
 <7.5 MET-hours/week 376 (34.9%) 28 (31.1%) 0.64 
 ≥7.5 MET-hours/week 680 (63.2%) 61 (67.8%)  
 Unknown 20 (1.9%) 1 (1.1%)  
Calcium intake 
 Tertile 1 333 (30.9%) 31 (34.4%) 0.50 
 Tertile 2 346 (32.2%) 33 (36.7%)  
 Tertile 3 345 (32.1%) 23 (25.6%)  
 Unknown 52 (4.8%) 3 (3.3%)  
p53 antibodies
Seron = 1,076Sero+n = 90χ2P value
Median (range) age at blood draw (years) 71 (56–87) 72 (56–87)  
Median (range) years before diagnosis 4.1 (<1–10.8) years 3.4 (<1–9.8 years)  
Sex 
 Female 553 (51.4%) 51 (56.7%) 0.34 
 Male 523 (48.6%) 39 (43.3%)  
Race 
 White 1,057 (98.2%) 88 (97.8%) 0.93 
 Black 8 (0.7%) 1 (1.1%)  
 Other/unknown 11 (1%) 1 (1.1%)  
Education 
 Less than high school 36 (3.3%) 3 (3.3%) 0.57 
 High school/missing education 230 (21.4%) 25 (27.8%)  
 Some college/vocational school 340 (31.6%) 26 (28.9%)  
 College/graduate school 470 (43.7%) 36 (40.0%)  
BMI 
 <25.0 454 (42.2%) 45 (50.0%) 0.10 
 25.0–<30.0 431 (40.1%) 27 (30.0%)  
  ≥30.0 172 (16.0%) 14 (15.6%)  
 Unknown 19 (1.8%) 4 (4.4%)  
Smoking status 
 Never 485 (45.1%) 37 (41.1%) 0.60 
 Current 28 (2.6%) 4 (4.4%)  
 Former 549 (51.0%) 47 (52.2%)  
 Unknown 14 (1.3%) 2 (2.2%)  
Alcohol intake 
 Nondrinker 261 (24.3%) 24 (26.7%) 0.76 
 ≤1 drink/day 563 (52.3%) 44 (48.9%)  
 >1 drink/day 244 (22.7%) 22 (24.4%)  
 Unknown 8 (0.7%) 0 (0%)  
Physical activity 
 <7.5 MET-hours/week 376 (34.9%) 28 (31.1%) 0.64 
 ≥7.5 MET-hours/week 680 (63.2%) 61 (67.8%)  
 Unknown 20 (1.9%) 1 (1.1%)  
Calcium intake 
 Tertile 1 333 (30.9%) 31 (34.4%) 0.50 
 Tertile 2 346 (32.2%) 33 (36.7%)  
 Tertile 3 345 (32.1%) 23 (25.6%)  
 Unknown 52 (4.8%) 3 (3.3%)  

Participants who were seropositive for p53 antibodies were more likely to subsequently develop colorectal cancer (RR = 1.77; 95% CI, 1.12–2.78; Table 2). Although the RR increased slightly with higher antibody levels, the CIs overlapped, and there was no evidence of statistical trend (Ptrend = 0.61). When results were stratified by time to diagnosis (Table 3), the association was strongest within 3 years (RR = 2.26; 95% CI, 1.06–4.83), but still suggestive within 6 years. After 6 years, there was no association between prediagnostic serum p53 antibodies and risk of colorectal cancer. The association between elevated p53 antibodies and subsequent colorectal cancer was most evident in individuals who had not had a colorectal endoscopy within 2 years before blood draw: RRseropositive = 2.66 (95% CI, 1.28–5.56). Clinical characteristics of the colorectal cancer cases are shown in Table 4. A higher percentage of seropositive cases had rectal cancer, distant stage, and died from their colorectal cancer, but none of these differences were statistically significant.

Table 2.

Relative risk of colorectal cancer for prediagnostic p53 antibodies by serostatus and antibody level

p53 antibody serostatusCases (n)Controls (n)RR (95% CI)a
All participants 
 Seronegative 351 725 1.00 
 Seropositive 41 49 1.77 (1.12–2.78) 
  Low (tertile 1) p53 antibody level 13 17 1.61 (0.77–3.39) 
  Medium (tertile 2) p53 antibody level 14 16 1.86 (0.88–3.92) 
  High (tertile 3) p53 antibody level 14 16 1.84 (0.85–4.00) 
Participants who had not received a colorectal endoscopy within 2 years before p53 antibody measurement 
 Seronegative 208 297 1.00 
 Seropositive 24 15 2.66 (1.28–5.56) 
p53 antibody serostatusCases (n)Controls (n)RR (95% CI)a
All participants 
 Seronegative 351 725 1.00 
 Seropositive 41 49 1.77 (1.12–2.78) 
  Low (tertile 1) p53 antibody level 13 17 1.61 (0.77–3.39) 
  Medium (tertile 2) p53 antibody level 14 16 1.86 (0.88–3.92) 
  High (tertile 3) p53 antibody level 14 16 1.84 (0.85–4.00) 
Participants who had not received a colorectal endoscopy within 2 years before p53 antibody measurement 
 Seronegative 208 297 1.00 
 Seropositive 24 15 2.66 (1.28–5.56) 

aControlling for matching factors and time between blood draw and colorectal cancer diagnosis.

Table 3.

Relative risk of colorectal cancer for prediagnostic p53 antibodies by serostatus stratified on years before diagnosis

Years between diagnosis (or matched case's diagnosis) and blood draw
<3 years3–<6 years6+ years
p53 antibody serostatusCasesControlsRRa (95% CI)CasesControlsRRa (95% CI)CasesControlsRRa (95% CI)CasesControlsRR a (95% CI)
Seronegative 351 725 1.00 125 260 1.00 116 244 1.00 110 221 1.00 
Seropositive 41 49 1.77 (1.12–2.78) 18 18 2.26 (1.06–4.83) 15 17 1.84 (0.89–3.79) 14 1.15 (0.44–2.99) 
Years between diagnosis (or matched case's diagnosis) and blood draw
<3 years3–<6 years6+ years
p53 antibody serostatusCasesControlsRRa (95% CI)CasesControlsRRa (95% CI)CasesControlsRRa (95% CI)CasesControlsRR a (95% CI)
Seronegative 351 725 1.00 125 260 1.00 116 244 1.00 110 221 1.00 
Seropositive 41 49 1.77 (1.12–2.78) 18 18 2.26 (1.06–4.83) 15 17 1.84 (0.89–3.79) 14 1.15 (0.44–2.99) 

aControlling for matching factors and time between blood draw and colorectal cancer diagnosis.

Table 4.

Clinical characteristics of colorectal cancer cases by p53 serostatus <1–11 years before diagnosis

CharacteristicsSero (n = 329)Sero+ (n = 40)Fisher exact P valuea
Subsite 
 Colon 267 (81.2%) 30 (75%) 0.40 
 Rectum 62 (18.8%) 10 (25%)  
Stage 
 Localized 137 (41.6%) 16 (40%)  
 Regional 150 (45.6%) 15 (37.5%) 0.18 
 Distant  33 (10%)  8 (20%)  
 Missing stage  9 (2.7%)  1 (2.5%)  
Grade 
 Well differentiated  33 (10%)  8 (20%)  
 Moderately differentiated 188 (57.1%) 17 (42.5%) 0.09 
 Poorly differentiated  65 (19.8%) 10 (25%)  
 Undifferentiated  9 (2.7%)  2 (5%)  
 Unknown  34 (10.3%)  3 (7.5%)  
Died of colorectal cancer* 
 No 266 (75.8%) 27 (65.9%) 0.18 
 Yes  85 (24.2%) 14 (34.1%)  
CharacteristicsSero (n = 329)Sero+ (n = 40)Fisher exact P valuea
Subsite 
 Colon 267 (81.2%) 30 (75%) 0.40 
 Rectum 62 (18.8%) 10 (25%)  
Stage 
 Localized 137 (41.6%) 16 (40%)  
 Regional 150 (45.6%) 15 (37.5%) 0.18 
 Distant  33 (10%)  8 (20%)  
 Missing stage  9 (2.7%)  1 (2.5%)  
Grade 
 Well differentiated  33 (10%)  8 (20%)  
 Moderately differentiated 188 (57.1%) 17 (42.5%) 0.09 
 Poorly differentiated  65 (19.8%) 10 (25%)  
 Undifferentiated  9 (2.7%)  2 (5%)  
 Unknown  34 (10.3%)  3 (7.5%)  
Died of colorectal cancer* 
 No 266 (75.8%) 27 (65.9%) 0.18 
 Yes  85 (24.2%) 14 (34.1%)  

*Cases identified through linkage to the National Death Index (and not verified through registry linkage) were only included in this table for number of deaths due to colorectal cancer because we do not have stage, grade, or subsite information for those cases (n = 23).

aMissing categories are not included in the calculated P value.

In this nested case–control study of p53 antibodies measured in blood samples collected up to 10.8 years before colorectal cancer diagnosis (minimum <1 year), we found that individuals who were seropositive for p53 antibodies had an approximately 2-fold higher risk of developing colorectal cancer within the next 3 years compared with individuals who were seronegative. The association was particularly apparent among individuals who had not had a recent colorectal endoscopy at the time of blood draw. One possible reason for this finding is that those who had been recently screened were less likely to have a colorectal cancer close to blood draw. Although we lack the sample size to fully understand the relevant time window for this association, our results suggest that the association is strongest close to blood draw. Another possible reason for the stronger observed association among unscreened participants is that elevated p53 antibody serostatus in some screened participants may reflect a cancer that was in progress but was removed at the polyp stage before it could fully develop. This idea is supported by the higher proportion of false positives among the screened participants (67%) than among the unscreened (33%). To our knowledge, this is the largest study of prediagnostic p53 antibodies and risk of colorectal cancer to date, with almost four times more cases than the only other published study on the topic (3). Unlike the previous study, we could examine clinical characteristics of the cases by p53 antibody serostatus. Of note, p53 seropositive cases were observed across all stages, grades, and subsites. However, a higher proportion of seropositive cases were diagnosed at distant stage. This finding was based on 41 total distant stage cancers (8 seropositive), however, and needs further study. Our analyses were limited in that we could not identify the exact timing of when participants seroconverted to p53 seropositivity as we did not have serial blood measurements. However, our stratified results suggest that the association between p53 antibodies and colorectal cancer risk is strongest within the first 3 years of blood draw. These results are consistent with those of Pederson and colleagues (3) and indicate that elevated p53 antibodies can be detected before diagnosis of colorectal cancer. The precise timing of the appearance of p53 autoantibodies should be explored in future studies.

Of note, only 11% of colorectal cancer cases in our analysis were seropositive for p53 antibodies at blood draw. Autoantibodies to p53 do not appear to have sufficient diagnostic sensitivity to serve as a single early detection marker, which is consistent with other known blood markers like carcinoembryonic antigen (CEA; ref. 11). In a review of blood markers measured after colorectal cancer diagnosis (1), CEA (examined in 4 studies) had slightly poorer specificity (89%–96%), but slightly better sensitivity (9%–64%), compared with p53 autoantibodies (examined in 21 studies), which had specificity ranging from 90% to 100% and sensitivity from 9% to 46%.

By itself, measurement of p53 antibodies is not sufficiently informative to be clinically useful, but it may be a good candidate to use in combination with other blood markers such as CEA, c-myc, Annexin A4, and Mucin 1. These markers along with several others have been detected in colorectal cancer patients by at least one previous study (1). Blood markers, used along with lifestyle and other information, may help clinicians target screening at individuals who are at higher risk of having an undetected colorectal cancer, or of developing a colorectal cancer in the near future.

No potential conflicts of interest were disclosed.

Conception and design: L.R. Teras, S.M. Gapstur, M. Pawlita, P.T. Campbell

Development of methodology: L.R. Teras, M. Pawlita, T. Waterboer, P.T. Campbell

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S.M. Gapstur, E.J. Jacobs, M. Pawlita, T. Waterboer, P.T. Campbell

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L.R. Teras, S.M. Gapstur, M.L. Maliniak, E.J. Jacobs, A. Michel, M. Pawlita, T. Waterboer, P.T. Campbell

Writing, review, and/or revision of the manuscript: L.R. Teras, S.M. Gapstur, M.L. Maliniak, E.J. Jacobs, T. Gansler, A. Michel, M. Pawlita, T. Waterboer, P.T. Campbell

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): L.R. Teras, M.L. Maliniak, A. Michel

Study supervision: M. Pawlita, P.T. Campbell

The American Cancer Society funds the creation, maintenance, and updating of the Cancer Prevention Study-II (CPS-II) cohort. The authors thank the CPS-II participants and Study Management Group for their invaluable contributions to this research. The authors would also like to acknowledge the contribution to this study from central cancer registries supported through the Centers for Disease Control and Prevention National Program of Cancer Registries, and cancer registries supported by the National Cancer Institute Surveillance Epidemiology and End Results program.

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.

1.
Chen
H
,
Werner
S
,
Tao
S
,
Zörnig
I
,
Brenner
H
. 
Blood autoantibodies against tumor-associated antigens as biomarkers in early detection of colorectal cancer
.
Cancer Lett
2014
;
346
:
178
87
.
2.
Hundt
S
,
Haug
U
,
Brenner
H
. 
Blood markers for early detection of colorectal cancer: a systematic review
.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
1935
53
.
3.
Pedersen
J
,
Gentry-Maharaj
A
,
Fourkala
E
,
Dawnay
A
,
Burnell
M
,
Zaikin
A
, et al
Early detection of cancer in the general population: a blinded case–control study of p53 autoantibodies in colorectal cancer
.
Br J Cancer
2013
;
108
:
107
14
.
4.
Suppiah
A
,
Greenman
J
. 
Clinical utility of anti-p53 auto-antibody: systematic review and focus on colorectal cancer
.
World J Gastroenterol
2013
;
19
:
4651
.
5.
Soussi
T
. 
p53 Antibodies in the sera of patients with various types of cancer: a review
.
Cancer Res
2000
;
60
:
1777
88
.
6.
The Cancer Genome Atlas Research Network
. 
Comprehensive molecular characterization of human colon and rectal cancer
.
Nature
2012
;
487
:
330
7
.
7.
Calle
EE
,
Rodriguez
C
,
Jacobs
EJ
,
Almon
ML
,
Chao
A
,
McCullough
ML
, et al
The American cancer society cancer prevention study II nutrition cohort: rationale, study design, and baseline characteristics
.
Cancer
2002
;
94
:
2490
501
.
8.
Waterboer
T
,
Sehr
P
,
Michael
KM
,
Franceschi
S
,
Nieland
JD
,
Joos
TO
, et al
Multiplex human papillomavirus serology based on in situ-purified glutathione s-transferase fusion proteins
.
Clin Chem
2005
;
51
:
1845
53
.
9.
Kjaerheim
K
,
Roe
OD
,
Waterboer
T
,
Sehr
P
,
Rizk
R
,
Dai
HY
, et al
Absence of SV40 antibodies or DNA fragments in prediagnostic mesothelioma serum samples
.
Int J Cancer
2007
;
120
:
2459
65
.
10.
Sehr
P
,
Zumbach
K
,
Pawlita
M
. 
A generic capture ELISA for recombinant proteins fused to glutathione S-transferase: validation for HPV serology
.
J Immunol Methods
2001
;
253
:
153
62
.
11.
Nielsen
HJ
,
Jakobsen
KV
,
Christensen
IJ
,
Brünner
N
. 
Screening for colorectal cancer: possible improvements by risk assessment evaluation?
Scandinavian J Gastroenterol
2011
;
46
:
1283
94
.