[Gann, 75, 130-135; February, 1984] HIGH CONCENTRATIONS OF N-TERMINAL PEPTIDE OF TYPE III PROCOLLAGEN IN THE SERA OF PATIENTS WITH VARIOUS CANCERS, WITH SPECIAL REFERENCE TO LIVER CANCER Terumasa HATAHARA, Shogo IGARASHI and Naoya FUNAKI The Department of Internal Medicine, Tokyo Hospital, Japan Tobacco and Salt Public Corporation, 1-4-3, Mita, Minato-ku, Tokyo 113 The concentrations of N-terminal peptide of type III procollagen in the sera of patients with various cancers were measured by radioimmunoassay. The mean value (with standard deviation) in the control group was 9.9±2.6 ng/ml. Serum levels exceeding 15 ng/ml were defined as positive, and it was found that 94% of 18 patients with primary liver cancer with cirrhosis, 88% of 8 patients with primary liver cancer without cirrhosis, 77% of 13 patients with metastatic liver cancer, 86% of 7 patients with recurrent breast cancer, 86% of 8 patients with colonic cancer, 75% of 8 patients with pancreatic cancer, 70% of 23 patients with stomach cancer, 51% of 35 patients with lung cancer, and 54% of 28 patients with uterine cancer showed positive levels. The concentrations showed great intersubject variations, probably reflecting the activity of tumor growth and/or invasion. The concentrations in the sera of patients with primary liver cancer with cirrhosis were generally higher than those in patients with liver cirrhosis alone or primary liver cancer without cirrhosis. This result suggested that the growth of primary liver cancer complicated by cirrhosis might be detected by serial measurements of this peptide in the serum of patients with liver cirrhosis. Present data suggested that this peptide is not cancer-specific, but assay of the peptide might be of value as an auxiliary means of detecting and monitoring various cancers, especially liver cancer. Key words: Collagen synthesis - N-Terminal peptide of type III procollagen - Tumor marker - Liver cancer - Liver cirrhosis Collagen is a representative of a family of structurally related proteins composing connective tissue and, in terms of weight, accounting for one-third of total body proteins. At least five phenotypically distinct types of collagen, i.e., type I through type V, have now been identified. The metabolism and anatomical localization of different types of collagen in tissues under normal, pathological and developmental conditions have been investigated.', I" Procollagen, the precursor of collagen, has additional peptides on both ends and these are referred to as N-terminal peptide and C-terminal peptide, respectively. In the metabolic conversion of procollagen to collagen, these additional peptides are removed and liberated into the circulation.1, ' The N-terminal peptide of type III procollagen derived from fetal calf was extensively investigated and a radioimmunoassay system was developed by Timpl and Glanville.10I An assay system employable in man was also developed.' 81 The concentration of this peptide in human serum specifically reflects the amount of type III collagen which has been newly synthesized in tissues, but does not depend on age or on sex from over 20 years of age.4.7.8' Several 130 Gann
ASSAY OF TYPE III PROCOLLAGEN IN SERUM workers demonstrated that the serum concentration of this peptide reflects the activity of collagen biosynthesis in various diseases, especially those involving the liver.2,1,11 Our preliminary study showed that high concentrations of this peptide were detected in the sera of patients with liver cancer.') As a next step, this peptide was assayed in the sera of patients with various cancers, including liver cancer, since it is possible that collagen synthesis may be proportionally enhanced in the presence of tumor growth or invasion, associated with increased interstitial tissue components. MATERIALS AND METHODS The patients, aged from 22 to 79 years old, consisted of 175 with various cancers and 146 with various liver diseases, who had been admitted to The Tokyo Hospital, Japan Tobacco and Salt Public Corporation, during the period from January 1978 to February 1983. In addition, 20 patients with rheumatoid arthritis and 5 with scleroderma, aged from 25 to 53 years old, who had visited The Department of Medicine and Physical Therapy, Faculty of Medicine, University of Tokyo, during the period from March 1977 to July 1982, were included. As a control, 15 normal subjects, aged from 33 to 58 years old, who had visited The Tokyo Hospital for health checks, were selected. The patients with various cancers consisted of 39 with liver cancer (primary liver cancer with cirrhosis, 18; primary liver cancer without cirrhosis, 8; metastatic liver cancer, 13), 23 with stomach cancer, 35 with lung cancer, 28 with uterine cancer, 2 with ovarian cancer, 5 with esophageal cancer, 8 with pancreatic cancer, 7 with colonic cancer, 2 with sigmoid and rectal cancer, 5 with lymphoma, 17 with breast cancer (including 7 recurrent cases), 3 with embryonal cancer, and 2 with gall bladder cancer. The patients with various liver diseases consisted of 35 with acute hepatitis, 52 with chronic hepatitis, 21 with alcoholic liver disease and 38 with liver cirrhosis. The sera were obtained from the patients before any treatment and from normal subjects in a fasting state, and were stored at -40 until assay. Diagnoses were based on clinical, biochemical, serological, roentogenographic, and/or histological criteria. The assay of N-terminal peptide of type III procollagen was performed according to the method described by Rohde et al.8) Radioimmunoassay kits purchased from Behringwerke, Marburg, Germany, were employed. The kits comprise '25I-procollagen-III-peptide(freeze-dried, 200 ng: Ca, 10 Ci), procollagen-iii-peptide antiserum (raised in rabbits, freeze-dried), standard procollagen-iii-peptide (freeze-dried, 5.26 ng/ml), separating reagent (freeze-dried, anti-rabbitglobulin) and buffer tablets (phosphate-buffered saline). Six concentrations of standard procollagen (5.26, 2.63, 1.32, 0.66, 0.33 and 0.17 ng/ml) and four different dilutions of serum sample (1/5, 1/10, 1/20 and 1/40), 200 p1, were prepared and 100 pl of procollagen-iii-peptide antiserum was added to each tube, then the mixture was incubated for 16 to 24 hr at 4. Thereafter, 100 p1 of 725I-procollagen-III-peptide was added, and the tubes were incubated at 4 for 7 hr. Finally, 500 pl of separating reagent was added and incubation was continued at 4 for 16 to 24 hr. After centrifugation at 2,000 rpm for 20 min, the supernatant was carefully removed by decantation. All experiments were done in duplicate. Radioactivity of precipitated antigen-antibody complex was measured in a gamma scintillation counter and the concentration of the peptide was calculated by the 50% intercept method. RESULTS The Concentrations of N-Terminal Peptide of Type III Procollagen in the Sera of Normal Subjects and Patients with Various Liver Diseases, Rheumatoid Arthritis and Scleroderma Fig. 1 shows the ranges of serum concentrations of this peptide among the groups of normal subjects, various liver diseases, rheumatoid arthritis and scleroderma. The sera of the normal subjects showed concentrations of this peptide ranging from 6.3 to 15.8 ng/ml with a mean value and standard deviation of 9.9+ 2.6 ng/ml. The mean values of this peptide in the sera of patients with acute hepatitis, chronic hepatitis, liver cirrhosis and liver cancer were 28.3±29.6, 19.3±17.3, 21.4 ±24.6 and 77.0+85.9 ng/ml, respectively. The sera of patients with so-called collagen diseases, rheumatoid arthritis and scleroderma, showed mean levels of 9.9+3.9 and 7.4+3.1 ng/ml, respectively. 75(2) 1984 131
T. HATAHARA, ET AL. Fig. 1. The concentrations of N-terminal peptide of type III procollagen in the sera of normal subjects and patients with various liver diseases, rheumatoid arthritis and scleroderma Concentrations exceeding 100 ng/ml are indicated with the assay value. P-III-P means the N-terminal peptide of type III procollagen. The solid line shows the mean value for normal subjects and the broken line shows a level of 15 ng/ml, which was defined as the upper limit among normal subjects. The Concentrations of N-Terminal Peptide of Type III Procollagen in the Sera of Patients with Various Cancers Fig. 2 shows the ranges of serum concentrations of this peptide among the various cancers. Liver cancer was classified into two groups, primary and metastatic. The sera of patients with primary liver cancer and metastatic liver cancer showed mean values of 80.3± 93.6 and 72.1±71.3 ng/ml, respectively. The mean concentrations in the sera of patients with stomach, lung, uterine, pancreatic, and colonic cancers were 71.2± 180.1, 22.1 + 20.1, 99.9 + 270.4, 129.2 + 201.1, and 82.3±117.4 ng/ml, respectively. The mean value of this peptide concentration in the sera of 17 patients with breast cancer was 189.4±576.1 ng/ml, while that of 7 recurrent cases was 895.8+885.5 ng/ml. 132 Gann
ASSAY OF TYPE III PROCOLLAGEN IN SERUM Fig. 2. The concentrations of N-terminal peptide of type III procollagen in the sera of patients with various cancers Concentrations exceeding 100 ng/ml are indicated with the assay value. P-III-P means the N-terminal peptide of type III procollagen. The solid line shows the mean value for normal subjects and the broken line shows a level of 15 ng/ml, which was defined as the upper limit among normal subjects. Liver Cancer with or without Cirrhosis Liver cancer (primary or metastatic) was classified into two groups, i.e., with and without cirrhosis. Fig. 3 shows the concentration ranges of this peptide in the sera of patients with primary liver cancer with or without cirrhosis, metastatic liver cancer with or without cirrhosis and liver cirrhosis alone. The mean values in the sera of patients with primary liver cancer with cirrhosis, that without cirrhosis, metastatic liver cancer and liver cirrhosis alone were 44.4 ±44.9, 96.3±105.8, 72.1± 71.3, and 21.4 ±24.6 ng/ml, respectively. 75(2) 1984 133
T. HATAHARA, ET AL. 54% of 28 patients with uterine cancer, 75% of 8 patients with pancreatic cancer, 86% of 7 patients with colonic cancer, and 65% of 17 patients with breast cancer (80% of 7 patients with recurrent cases) showed positive levels. Because of the limited numbers of patients with ovarian, esophageal, gall bladder, and embryonal cancers, the results were not conclusive in these cases. Fig. 3. The concentrations of N-terminal peptide of type III procollagen in the sera of patients having primary liver cancer, liver cirrhosis alone, and metastatic liver cancer 0, liver cancer without cirrhosis;, liver cancer with cirrhosis. Concentrations exceeding 100 ng/ml are indicated with the assay value. P-III-P means the N- terminal peptide of type III procollagen. The solid line shows the mean value for normal subjects and the broken line shows a level of 15 ng/ml, which was defined as the upper limit among normal subjects. Occurrence of N-Terminal Peptide of Type III Procollagen at Concentrations Exceeding 15 ng/ml Based on the range (from 6.3 to 15.8 ng/ml) and mean value (9.9±2.6 ng/ml) of the normal subjects, serum levels exceeding 15 ng/ml were tentatively defined as positive. Based on this criterion, 94% of 18 patients with primary liver cancer with cirrhosis, 88% of 8 patients with primary liver cancer without cirrhosis, 77% of 13 patients with metastatic liver cancer, 70% of 23 patients with stomach cancer, 51 % of 35 patients with lung cancer, DISCUSSION For the clinical evaluation of disease activity in relation to collagen synthesis, the assay of several enzyme activities, such as collagenase or prolylhydroxylase, in serum and tissue and the assay of urinary hydroxyproline have been attempted. However, these assays are neither adequately sensitive nor collagen type-specific. On the other hand, the assay of N-terminal peptide of type III procollagen is more sensitive, is specific to type III collagen and is independent of age or of sex at over 20 years of age, so the quantification of this peptide in serum could be of clinical significance for the evaluation of disease activity in terms of collagen biosynthesis. We reported high concentrations of this peptide in the sera of patients with liver cancer. 21 Moreover, as shown in Fig. 2, not only liver cancer but also other types of cancers seem to affect collagen biosynthesis. The concentrations of this peptide in the sera of patients with various cancers showed great intersubject variations, most likely reflecting the activity tumor growth and/or invasion. The present results suggest that the assay of this peptide in serum is not cancer-specific, and thus is not useful for differential diagnosis, but it might be of value as an auxiliary means of detecting various cancers in combination with other diagnostic methods. As shown in Fig. 3, the concentrations in the sera of patients with primary liver cancer with cirrhosis were generally higher than those in patients with liver cirrhosis alone or primary liver cancer without cirrhosis. 134 Gann
ASSAY OF TYPE III PROCOLLAGEN IN SERUM This result means the collagen synthesis was less active in liver cirrhosis alone, while it was enhanced when cancer and cirrhosis were present together. In other words, serial measurement of this peptide in the sera of patients with liver cirrhosis might be useful for the detection of hepatocellular carcinoma, which often complicates liver cirrhosis. As regards the effect of treatment, such as X-ray irradiation or chemotherapeutics, the serum of one patient with liver cancer showed a concentration of 400 ng/ml at the time of admission, and a concentration of 90 ng/ml after several months of chemotherapy, but more data are needed before such measurements can be used for clinical evaluation of the effectiveness of treatment. An interesting question is what kind of cell or tissue produces type III collagen actively under the condition of tumor growth and/or invasion. Recent evidence has indicated that various cells in addition to fibroblasts have the potential to produce collagen.', 3, 9> Although fibroblasts are probably responsible for the increase of this peptide concentration in serum, the possibility that the tumor cell itself produces collagen cannot be ruled out. In conclusion, our present results demonstrate the presence of high concentrations of N-terminal peptide of type III procollagen in the sera of patients with various cancers, suggesting that the measurement of serum concentration of this peptide may serve as a means for detecting various cancers, especially liver cancer accompanying cirrhosis. ACKNOWLEDGMENTS The authors are grateful to Drs. Abe and Kimura, National Cancer Center, and Dr. Shibata, Hoechst Japan Ltd. for fruitful discussions and Mr. Star, Mr. Yashima and Ms. Kameyama, Hoechst Japan Ltd. for help in preparing the manuscript and surveying clinical data, respectively. The sera of patients with rheumatoid arthritis and scleroderma were kindly provided by Drs. Hirose and Inoue, The Department of Medicine and Physical Therapy, The Faculty of Medicine, University of Tokyo. (Received Aug. 17, 1983/Accepted Nov. 21, 1983) REFERENCES 1) Borstein, P. Structurally distinct collagen types. Ann. Rev. Biochem., 49, 957-1003 (1980). 2) Hatahara, T., Funaki, N. and Igarashi, S. The measurement of amino-terminal procollagen peptide in liver diseases. Kanzo (Acta Hepatol. Jpn.), 24, 78-78 (1983) (in Japanese). 3) Hatahara, T. and Seyer, J. M. Procollagen production by rat hepatocytes in primary culture. Biochim. Biophys. Acta, 716, 431-438 (1982). 4) Nakano, H., Kawasaki, T., Miyanuma, M., Fukuda, Y. and Imura, H. The concentration of amino-terminal peptide of type III procollagen related to age and liver fibrosis. Kanzo (Acta Hepatol. Jpn.), 24, 796-796 (1983) (in Japanese). 5) Prockop, D. J., Kivirikko, K. I. and Tuderman, L. The biosynthesis of collagen and its disorder. New Eng. J. Med., 301, 13-23 (1979). 6) Prockop, D. J., Kivirikko, K. I. and Tuderman, L. The biosynthesis of collagen and its disorder. New Eng. J. Med., 301, 77-85 (1979). 7) Rohde, H., Hahn, E. and Timpl, R. Radioimmunoassay of amino-terminal procollagen peptide in liver disease. Fresenius Z. Anal. Chem., 290, 151-152 (1978). 8) Rohde, H., Vargas, L., Hahn, E., Kalbfleisch, H., Bruguera, M. and Timpl, R. Radioimmunoassay for type III procollagen peptide and its application to human liver disease. Eur. J. Clin. Invest., 9, 451-459 (1979). 9) Sakakibara, K. and Umeda, M. Production of collagen and acidic glycosaminoglycans by an epithelial liver clone in culture. Exp. Cell Res., 110, 159-165 (1977). 10) Timpl, R. and Glanville, R. W. The aminopropeptide of collagen. Clin. Orthop. Related Res., 158, 224-242 (1981). 11) Von Der Mark, K. Localization of collagen types in tissue. In Int. Rev. Connect. Tissue Res., 9, 263-313 (1981). 75(2) 1984 135