Fax: 011 (81) 0432262113
Titration of serum p53 antibodies in 1085 patients with various types of malignant tumors
A multiinstitutional analysis by the Japan p53 antibody research group
Version of Record online: 17 JAN 2003
Copyright © 2003 American Cancer Society
Volume 97, Issue 3, pages 682–689, 1 February 2003
How to Cite
Shimada, H., Ochiai, T. and Nomura, F. (2003), Titration of serum p53 antibodies in 1085 patients with various types of malignant tumors. Cancer, 97: 682–689. doi: 10.1002/cncr.11092
- Issue online: 17 JAN 2003
- Version of Record online: 17 JAN 2003
- Manuscript Accepted: 5 SEP 2002
- Manuscript Revised: 21 AUG 2002
- Manuscript Received: 17 JUN 2002
- Ministry of Education, Science, and Culture of Japan. Grant Number: 12671200
- serum p53 antibody;
- tumor marker;
- carcinoembryonic antigen
There have been very few large-scale, multiinstitutional studies of surveillance of serum p53 antibodies (S-p53 Abs) in patients with various malignant tumors.
A highly specific, quantitative enzyme-linked immunosorbent assay (ELISA) kit was developed and used to evaluate the efficiency of detecting p53 Abs. A cut-off value was established by analyzing sera from 205 healthy volunteers as reference individuals. Sera from 1085 patients with various types of primary malignant tumors were studied for the presence of S-p53 Abs before treatment. Sera from 34 patients were selected randomly for a competition assay to ensure that antibodies were specific to p53 protein. Carcinoembryonic antigen (CEA) was assessed to compare its positive rate with the positive rate of S-p53 Abs.
The median value of S-p53 Abs in healthy control individuals was 0.33 U/mL (range, 0.0–4.39 U/mL). Based on reference values that were calculated using parametric determination of the lower 0.95 fraction of the reference distribution in healthy control individuals, the cut-off value was determined as 1.3 U/mL. Two hundred twenty-one of 1085 patients (20.4%) were positive for S-p53 Abs. The highest relevance of S-53 Abs was associated with head and neck carcinoma (32%), followed by esophageal carcinoma (30%), colorectal carcinoma (24%), and carcinoma of the uterus (23%). The positive rate for S-p53 Abs was higher compared with the positive rate for CEA in patients with squamous cell carcinoma.
Surveillance of S-p53 Abs is useful in detecting various types of malignant tumors, particular in patients with squamous cell carcinoma. Cancer 2003;97:682–9. © 2003 American Cancer Society.
Currently available serologic tumor markers are not satisfactory in diagnosing malignancies, because these serum proteins are not sensitive and/or specific for carcinoma. Although carcinoembryonic antigen (CEA) is the most popular tumor marker, it has been reported that CEA levels are increased in patients who smoke or who have inflammation.1–3 Studies of the molecular biology of malignant tumors have emphasized the importance of a number of protooncogenes and tumor suppressor genes in human malignancies. Thus, the search for biomarkers that can diagnose various types of malignancies is important for the better management of patients.
Several studies have reported that serum p53 antibodies (S-p53 Abs) were detected previously in different populations that were at increased risk for developing malignant disease.4–6 S-p53 Abs can be used to follow the response of patients with malignant tumors during treatment.7 The presence of S-p53 Abs has been correlated with a poor prognosis for patients with several tumor types.8–10
We previously reported a high prevalence of S-p53 Abs in patients with esophageal carcinoma and colorectal carcinoma11, 12 using the first version of the p53 Abs enzyme-linked immunosorbent assay (ELISA) kit (PharmaCell, Paris, France). Recently, the second version of a highly specific, quantitative p53 Abs ELISA kit (MESACUP anti-p53 Test; Medical & Biological Laboratories Co. Ltd. (MBL); Nagoya, Japan), has been developed. We took advantage of this novel assay to evaluate diagnostic values of S-p53 Abs in a large spectrum of patients with various malignant tumors as part of a multiinstitutional study. Because data on sample storage conditions and an appropriate cut-off value are essential to conduct a multiinstitutional study, analytic validation of the assay system also was performed. Therefore, the objectives of this study were 1) to evaluate the stability of the titer of S-p53 Abs, 2) to determine the cut-off level of the titer of S-p53 Abs, 3) to determine the positive rate and the concentration of S-p53 Abs in each of the malignancies, 4) to compare the positive rate with the CEA rate, and 5) to compare the positive rate in patients who had squamous cell carcinoma with the positive rate in patients who had adenocarcinoma.
Although several studies for S-p53 Abs in patients with malignant disease have been performed, only a few have shown positive rates and cut-off values in a large-scale study. To our knowledge, this is the first multiinstitutional study to evaluate the significance of S-p53 Abs in patients with malignant tumors. We evaluated the positivity of S-p53 Abs in 1085 patients with 15 different types of malignant tumors from 20 cancer institutions in Japan.
MATERIALS AND METHODS
Patients and Controls
In total, 1085 patients with 15 types of malignant tumors were enrolled in this study. The patients with each tumor type included 301 patients with esophageal tumors, 192 patients with colorectal tumors, 125 patients with lung tumors, 123 patients with gastric tumors, 71 patients with breast tumors, 53 patients with tumors of the uterine cervix, 33 patients with bladder tumors, 31 patients with head and neck tumors, 31 patients with gliomas, 28 patients with pancreatic tumors, 27 patients with ovarian tumors, 23 patients with prostate tumors, 22 patients with uterine tumors, 19 patients with liver tumors, and 6 patients with biliary tract tumors. The study included 677 males (62.4%) and 408 females (37.6%), with a mean age of 61.8 years (range, 19–88 years). Patients were treated at 27 institutes of the Japan p53 Antibody Research Group between April, 2000 and April, 2001. Serum samples were obtained before treatment and stored at −80 °C until they were assayed.
Two hundred five healthy volunteers, including 91 males and 114 females with a mean age of 42.6 years (range, 24–82 years), were used as a control group to determine the reliability of seronegative testing for S-p53 Abs. In total, 189 patients with benign disease, including 72 males and 117 females with a mean age of 59.4 years (range, 21–86 years), were selected for the nontumor control group. The patients with each benign disease included 15 patients with benign brain tumors, 1 patient with a head and neck tumor, 12 patients with benign pulmonary disease, 5 patients with breast tumors, 2 patients with esophageal papillomas, 1 patient with gastric leiomyoma, 2 patients with tumor-forming pancreatitis, 1 patient with primary sclerosing cholangitis, 8 patients with ovarian tumors, 1 patient with myoma uteri, 14 patients with bladder disease, 13 patients with prostate hypertrophy, 21 patients with chronic hepatitis, 10 patients with inflammatory bowel disease, 40 patients with malignant rheumatoid arthritis, 9 patients with primary biliary cirrhosis, 21 patients with autoimmune hepatitis, and 13 patients with pemphigus. Patient recruitment and sample collections were performed within the guidelines of protocols approved by the Institutional Review Boards of the participating institutions. Informed consent was obtained from all participants.
Enzyme Immunoassay for S-p53 Abs
S-p53 Abs were assessed by ELISA with the anti-p53 EIA Kit II (MESACUP anti-p53 Test; MBL). In brief, the samples were added to the wells of a microtiter plate coated with wild type human p53 or a control protein and were incubated for 1 hour. A conjugated second antibody was added, and the wells were incubated for another hour; then, the substrate solution was added. After addition of the stop solution, the color reaction was measured immediately by the absorption at 450 nm using a photospectrometer. All samples were considered positive at an optical density greater than the density in the lowest positive control sample. Serum from a patient with lung carcinoma was diluted at various dilution factors and measured using the anti-p53 antibody ELISA kit. The serum samples that were diluted > 1:3 represented linearity, depending on the dilution factor. The antibody titer corresponding to the dilution factor 1:3 was defined as 15 U/mL.
A peroxidase-conjugated goat antihuman immunoglobulin G that binds anti-p53 Ab was added and incubated for another hour. The substrate solution was then added and incubated for 30 minutes. After addition of the stop solution, the color reaction was measured immediately by the absorption at 450 nm using a photospectrometer. A calibration curve was constructed from the specific signals of standards and from the levels of Abs indicated on the standard vials. This calibration curve is a linear regression curve that intersects the x-axis at 0. Levels of anti-p53 Abs were then determined from the calibration curve. The high-titer serum samples, which revealed > 15 U/mL, were diluted 1:10 and reevaluated. Seropositive patients were reevaluated 3 months after surgery.
Analytic Validation and Reference Values
Intraassay imprecision, expressed as the coefficient of variation (CV), was determined by analyzing three human serum samples containing 2.33 U/mL, 4.11 U/mL, and 9.88 U/mL of anti-p53 Abs in 8 parallel determinations. The interassay CV was determined by measuring each serum sample on 3 different days. Interference by hemolysis and hyperbilirubinemia was assessed by adding each potential interfering compound at various concentrations to serum samples of three different levels of anti-p53.
Two serum pools (at low and high anti-p53 levels) were prepared and stored in aliquots at − 80 °C, at 4 °C, and at room temperature. The influence of storage of ethylenediamine tetraacetic acid-whole blood prior to preparation of serum on anti-p53 Abs measurements also was tested at 4 °C or at 25 °C. Anti-p53 Ab levels were determined over 7 days. The frozen sera were subjected to indicate times of the freeze-thaw cycle. After these processes, the sera were measured with the anti-p53 Abs ELISA kit. Furthermore, the effects of repeated freezing and thawing of serum samples on anti-p53 Ab determination were assessed.
Two hundred five apparently healthy participants were recruited as a reference group according to the general a priori exclusion criteria suggested by IFCC documents on the theory of reference values.13 Reference values were calculated using the parametric determination of the lower 0.95 fraction of the reference distribution.
Sera from 34 of 1085 patients with various types of malignant tumors were selected randomly for a competition assay to ensure the presence of anti-p53 Abs. Those serum samples were diluted 1:100 using the sample dilution buffer with or without purified, wild type, full-length human p53 protein expressed by the baculovirus-insect cell expression system. Then, 100 μL of the solution were transferred to the wells of the anti-p53 Abs ELISA kit and measured according to the manufacturer's instructions. The amount of p53 protein in the solution was 100-fold greater compared with the amount of immobilized p53 protein in the well.
Assays for CEA
CEA concentrations were measured with the kit CEA-II EIA (Roche Diagnostics K.K., Tokyo, Japan). The cut-off value for serum CEA was 5.3 ng/mL, according to the manufacturer's instructions. Of 1085 patients, 603 patients (301 with squamous cell carcinoma and 302 with adenocarcinoma) were evaluated for CEA.
Fisher exact tests, Student t tests, and Mann–Whitney rank-sum tests were used to determine the significance between the two groups. A P value < 0.05 was considered significant.
Analytic Performances of the Anti-p53 ELISA Test
The lower limit of detection, defined as the concentration 3 standard deviations above the value for the level 0 calibrator, was 0.05 U/mL. The intraassay and interassay CVs were 1.85–2.37% and 0.30–2.32%, respectively, as presented in Table 1. Bilirubin levels up to 22.4 mg/dL and hemoglobin levels up to 450 mg/dL had no effect on the measurement of serum anti-p53 Abs (data not shown). No significant changes in anti-p53 Abs were observed when the whole blood samples were stored prior to serum preparation at − 80 °C or at room temperature for 7 days (Fig. 1). In addition, no significant changes in serum anti-p53 Abs levels were noted when the samples were stored at − 80 °C or at room temperature for 7 days (data not shown). Furthermore, repeated freezing and thawing (10 times) had no effect on the anti-p53 Abs measurements (data not shown).
|Anti-p53 Abs titer||Intraassay||Interassay|
|Mean ± SD (U/mL)||CV(%)||Mean ± SD (U/mL)||CV(%)|
|Low||2.33 ± 0.06||2.37||2.28 ± 0.04||1.98|
|Medium||4.11 ± 0.01||1.85||4.11 ± 0.01||0.30|
|High||9.88 ± 0.22||2.18||9.66 ± 0.22||2.32|
S-p53 Abs in Healthy Volunteers and Patients with Benign Disease
The median value of S-p53 Abs in healthy control donors was 0.33 U/mL (range, 0.0–4.39 U/mL). The median S-p53 Ab levels in 189 patients with benign disease was 0.31 U/mL (range, 0.0–16.9 U/mL). Based on reference values that were calculated using parametric determination of the lower 0.95 fraction of the reference distribution in healthy control donors, the cut-off value was determined as 1.3 U/mL.
Positivity Rate and Concentration of S-p53 Abs in Patients with Various Types of Malignant Tumors
Using 1.3 U/mL as the cut-off value, 221 of 1085 patients (20.4%) were positive for S-p53 Abs, and 10 of 205 healthy blood donors were positive for S-p53 Abs (specificity 95.5%). Thirteen of 189 patients (7%) with benign disease were positive for S-p53 Abs. The highest relevance of S-p53 Abs was associated with malignancies of the head and neck (35%), followed by esophageal carcinoma (30%), colorectal carcinoma (24%), uterine carcinoma (23%), and cervical carcinoma (19%) (Fig. 2).
S-p53 Ab concentrations were calculated for all samples, and the results are presented in Figure 3. The median value was 0.29 U/mL (range, 0–1716 U/mL). The highest titers were > 15 U/mL, which were seen in the sera from 51 patients. Among seropositive patients, 39 patients were reevaluated 1 month after surgery. Although 33 patients still revealed seropositivity, the mean value of S-p53 Abs was significantly lower compared with the value before surgery (data not shown).
Among 34 serum samples, S-p53 Abs titers in 19 samples were < 1.3 U/mL, and 15 samples were > 1.3 U/mL. All 15 serum samples with > 1.3 U/mL clearly showed inhibition of p53 Abs titers (Fig. 4). Thirteen of 19 serum samples with < 1.3 U/mL also showed inhibition of p53 Abs titers. Conversely, 6 of 19 serum samples did not show inhibition and were more likely to reveal less titer compared with the other serum samples. All 24 serum samples with titers > 0.94 U/mL showed inhibition of the p53 Abs titer.
Positivity of S-p53 Ab and CEA with Reference to Histologic Types
Six hundred three of 1085 patients (301 patients with squamous cell carcinoma and 302 patients adenocarcinoma) were evaluated for CEA levels. The overall positivity rate of S-p53 Abs (23.2%) was similar to that of CEA (25.5%). In patients with squamous cell carcinoma, the positivity rate of S-p53 Abs was higher compared with CEA (P = 0.05). In patients with adenocarcinoma, the positivity rate of CEA was higher compared with S-p53 Abs (P < 0.001). The positivity rate of S-p53 Abs in patients with squamous cell carcinoma was higher compared with that in patients with adenocarcinoma (P < 0.001) (Fig. 5).
The prevalence of S-p53 Abs in patients with malignant tumors has been studied occasionally; however, very few studies were reported as part of a multiinstitutional study. Several ELISA kits have been developed and have shown a comprehensive positivity rate compared with other conventional tumor markers.14–16 We investigated the positivity rate of S-p53 Abs in 1085 patients with 15 types of malignant tumors in this study using a new version of a quantitative ELISA kit. In this study, 20.4% of patients with malignant tumors were positive for S-p53 Abs. The positive rate of this ELISA kit was comparable to the rates found in previous studies of patients with carcinoma of the head and neck, esophageal carcinoma, colorectal carcinoma, uterine carcinoma, breast carcinoma, prostate carcinoma, and lung carcinoma.14–20
In the current study, we determined the cut-off value of S-p53 Abs by analyzing sera from 205 healthy control individuals. Previous studies revealed data from only a small number of healthy control participants.14, 15, 18 Our immunoassay had a specificity rate > 95.5%, which was comparable to the rtes found in other studies, using 1.3 U/mL as the cut-off value. In addition, a competition assay showed that all seropositive patients had Abs specific to p53 protein. Because all seropositive, healthy volunteers also revealed antibodies specific to p53 protein, intensive follow-up will be required to screen the subclinical potential of malignant tumors in the control group. Among patients with various benign diseases, 13 patients revealed positivity for S-p53 Abs. Most of the patients with benign disease showed a positive rate < 10%: The exceptions were patients with benign brain tumors and chronic hepatitis.
A surprising finding was that the anti-p53 levels in whole blood samples stored at room temperature were stable for 7 days. Because stability is the first consideration in screening outpatients, this ELISA kit may be useful. Furthermore, because the intraassay and interassay CVs were < 3%, this ELISA kit is satisfactory for further clinical use.
Among 15 types of tumors, a positive rate > 15% was revealed in the following types: carcinoma of the head and neck, esophageal carcinoma, colorectal carcinoma, uterine carcinoma, carcinoma of the uterine cervix, breast carcinoma, prostate carcinoma, and carcinoma of the biliary tract. In particular, carcinoma of the head and neck, esophageal carcinoma, colorectal carcinoma, and uterine carcinoma revealed > 20% positivity. Although squamous cell carcinoma antigen and CYFRA 21-1 are useful tumor markers in detecting tumors of the head and neck and the esophagus, S-p53 Abs also are useful for patients who test negative for conventional markers. We previously reported that S-p53 Abs were useful in detecting superficial esophageal carcinoma.11 In the current series, using the second version of the ELISA kit, we validated that positivity for S-p53 Abs in patients with superficial esophageal carcinoma and superficial colon carcinoma was higher greater compared with other conventional tumor markers (data not shown). Contrary to previous reports,21, 22 there were no patients with seropositive results among the 19 patients with hepatocellular carcinoma. In cervical carcinoma cells that contain high-risk human papilloma viruses, p53 is targeted for degradation by the human papilloma virus E6 oncoprotein.23 Because human papilloma virus was detected in 85% of patients with invasive cervical carcinoma, it was surprising that 19% of cervical carcinomas were positive for S-p53 Abs in the current study. Sobti et al. reported that none of their patients with cervical carcinoma had antibodies against the overexpressed p53.24 However, Numa et al. reported that S-p53 Abs were present in 12 of 86 patients (14%) with cervical carcinoma.25 Such discrepancies may be due in part to the differences in the ELISA systems used in those studies. Because the number of patients was small, another larger study will be required to finalize a conclusion about the significance of S-p53 Abs in the evaluation of patients with hepatocellular carcinoma and biliary tract carcinoma.
The positivity of S-p53 Abs alone was not sufficient for screening or monitoring of the patients with malignant tumors. Because the titer of S-p53 Abs was independent of other conventional tumor markers, such as CEA, tissue polypeptide antigen (TPA), squamous cell carcinoma antigen (SCC-Ag), and CYFRA21-1,26, 27 this assay can be used at least in association with those established tumor markers. In fact, among 301 patients with esophageal squamous cell carcinoma in this study, 16% of patients exhibited only S-p53 Abs, whereas serum values of the established tumor markers were lower than the cut-off levels (data not shown).
The positive rates of S-p53 Abs were compatible to the rates of p53 mutation in those malignant tumors.28 Positive correlations have been reported between p53 immunoreactivity and the presence of S-p53 Abs in patients with esophageal carcinoma,29 gastric carcinoma,30 colorectal carcinoma,31 and ovarian carcinoma.32 A strong correlation was reported between p53 mutation and the presence of S-p53 Abs.29, 33 We acknowledge the fact that p53 overexpression does not always indicate that a p53 mutation is present. It is well recognized that a significant proportion of p53 mutational damage is in the form of truncation type mutations, which typically result in minimal or completely absent immunoreactivity.
We were interested in the changes of S-p53 Abs titers with time and their relation to disease progression, regression, or therapeutic manipulation. Polge et al. reported that p53 antibody concentrations decreased rapidly 1 month after surgery in 6 of 10 patients with colorectal carcinoma.33 Zalcman et al. also reported that a correlation was found between the specific evaluation of p53 antibody titers and response to therapy. This suggests that p53 antibodies may represent a useful tool for checking the response to therapy and for monitoring some patients with recurrent tumors.34 In the current study, although 85% of patients with seropositive results still revealed seropositivity 1 month after surgery, the mean value of S-p53 Abs was significantly lower compared with the value before surgery (data not shown). Recurrent tumors developed only in patients who showed seropositivity 1 month after surgery. To date, none of the six patients who showed seroconversion 1 month after surgery developed recurrent tumors. These results claim the significance of S-p53 Abs monitoring in predicting recurrent tumors before they are detectable clinically. However, further intensive follow-up will be required to finalize a conclusion.
In summary, this study showed that 20.4% of malignant tumors could be detected by surveillance of S-p53 Abs with 95% specificity. The positivity rates in patients with carcinoma of the head and neck, esophageal carcinoma, colorectal carcinoma, and uterine carcinoma were > 20%. Because the ELISA assay is a quick and convenient assay for detecting p53 genetic alterations, S-p53 Abs may serve as a useful marker for routine screening in asymptomatic, high-risk patient groups. Additional work will be required to define the significance of monitoring after surgery and its prognostic impact.
This work was conducted as a multiinstitutional study by the Japan p53 Antibody Research Group. The authors thank the following members of the Japan p53 Antibody Research Group: Drs. M. Tsurumaru and Y. Kajiyama Y (Juntendo University School of Medicine); Drs. H. Makuuchi and H. Shimada (Tokai University School of Medicine); Drs. H. Kuwano and T. Asao (Gunma University School of Medicine); Drs. N. Kaminishi and N. Shimizu (Tokyo University School of Medicine); Drs. K. Miwa and I. Ninomiya (Kanazawa University Graduate School of Medical Science); Drs. H. Yamagishi and Y. Ueda (Kyoto Prefectural University of Medicine); Drs. K. Sugihara, K. Yoshinaga, S. Kishimoto, and N. Ishikawa (Tokyo Medical and Dental University); Drs. K. Hatakeyama, T. Suda, and T. Iiai (Niigata University School of Medicine); Dr. H. Monden (Osaka University Graduate School of Medicine); Drs. S. Matsuno and M. Sunamura (Tohoku University School of Medicine); Drs. H. Kato and H. Hashida (Hokkaido University Graduate School of Medicine); Dr. S. Imoto (National Cancer Center Hospital East); Drs. K. Kobayashi, M. Kawamura, and M. Gika (Keio University School of Medicine); Drs. A. Yoshihara, T. Oosaki, and K. Osato (Chiba Cancer Center); Dr. M. Kawata (Kawatetsu Chiba Hospital); Drs. N. Suzuki and H. Nakatsu (Asahi General Hospital); Dr. N. Ishige (Chiba National Hospital); and Drs. M. Miyazaki, K. Koda, K. Oda, H. Suzuki, T. Fujisawa, T. Iizasa, A. Konno, T. Numata, H. Muto, M. Sekiya, N. Tanaka, A. Yamaura, Y. Iwadate, H. Ito, T. Igarashi, H. Suzuki, H. Saisho, T. Tsuyuguchi, T. Ishihara, S. Kobayashi, S. Okazumi, Y. Nabeya, K. Saigo, S. Miyazaki, A. Takeda, H. Hayashi, T. Suzuki, Y. Miyazawa, and T. Ochiai (Chiba University Graduate School of Medicine).
- 13International Federation of Clinical Chemistry (IFCC). Scientific Committee, Clinical Section. Expert Panel on Theory of Reference Values (EPTRV). IFCC Document (1982) stage 2, draft 2, 1983-10-07 with a proposal for an IFCC recommendation. The theory of reference values. Part 2. Selection of individuals for the production of reference values. Clin Chim Acta. 1984; 139: 205F–213F.