Serum Heat Shock Protein 70 Levels and Lung Cancer Risk: A Case-Control Study Nested in a Large Cohort Study

Chronic inflammation has been hypothesized to play a role in the pathogenesis of several cancers (1). Heat shock proteins (Hsp) are molecular chaperones, mediating protein folding, transport, and formation of protein structures essential for cell survival under normal conditions (2). Hsp70 has been studied extensively and also seems to be the Hsp most inducible by stress. This protein is involved in inflammatory diseases, and serum levels of Hsp70 have been directly linked to the inflammatory status of subjects (3). C-reactive protein (CRP) is an acute-phase protein produced primarily in the liver in response to inflammation (4). Some studies have shown that serum CRP levels are associated with risk of coronary heart disease (5, 6). In recent years, elevated levels of CRP have been reported as a risk factor for the development of colon cancer (7). To the best of our knowledge, few epidemiologic studies have identified relationships between inflammatory markers and risk of lung cancer. We therefore examined correlations between risk of lung cancer and serum levels of both Hsp70 and high-sensitivity CRP (hsCRP) in a case-control study nested in a large-scale Japanese cohort.

Subjects were recruited in the Japan Collaborative Cohort (JACC) Study for Evaluation of Cancer Risk sponsored by Monbukagakusho (Ministry of Education, Culture, Sports, Science and Technology of Japan). This study involves 110,792 residents who were 40 to 79 years old at baseline from 45 areas all over Japan. Details of this study have been published elsewhere (8, 9).

An epidemiologic survey of lifestyle factors was conducted using a self-administered questionnaire about heath conditions and lifestyles, such as medical history, smoking habits, and alcohol consumption. For cigarette smoking, subjects were asked to describe their smoking status (never/former/current), age at starting smoking, average number of cigarettes smoked per day, years of smoking, and years of smoking cessation.

In addition to the questionnaire survey, participants in the JACC Study provided peripheral blood samples at health screening checkups sponsored by municipalities between 1988 and 1990. A total of 39,242 subjects (35.4% of respondents to the questionnaire survey) provided blood samples. Serum from each subject was divided into three to five tubes (100-500 μL/tube) and stored at −80°C until analyzed.

This study was approved by the Ethical Committee of Medical Care and Research at Fujita Health University.

Subjects who died or moved away from study areas were identified using population registries, and causes of death were confirmed from death certificates. Follow-up for death was conducted from baseline to the end of 1999, and only 4% of subjects were lost to follow-up due to moving during the study period. Death from lung cancer was defined by the code “C34” from the International Statistical Classifications of Diseases and Related Health Problems, 10th revision (10). During follow-up, 240 deaths from lung cancer were identified among subjects who had provided serum samples at baseline. Of these, 8 subjects with a history of lung and other cancers at baseline were excluded along with 23 subjects for whom suitable samples were lacking. For each case of lung cancer, 2 to 3 controls were selected from the remaining population without incident cancer or previous history of cancer, matching for sex, age (±3 years), and study area. Of the initial 569 matched controls, 68 subjects without sufficient samples for measurement of serum levels of both Hsp70 and hsCRP and 76 subjects whose matched cases removed from the study were excluded. Values for Hsp70 and hsCRP could not be measured in all serum samples, as some initial samples yielded insufficient sera and various other substances were also measured from the same samples. Finally, serum levels of hsCRP were measured in 209 cases and 425 controls in this study. Sufficient serum samples for determination of Hsp70 were available for 189 cases and 377 controls.

All samples were analyzed by trained staff blinded to case-control status in 2004. Serum Hsp70 and hsCRP were determined by ELISA using commercially available kits (Hsp70: Hsp70 ELISA kit, Stressgen Bioreagents, Victoria, BC Canada; hsCRP: hsCRP enzyme immunoassay test kit, Diagnostic Automation, Calabasas, CA) in our laboratory.

Body mass index (BMI) was calculated as body weight in kilograms divided by height in meters squared. Baseline characteristics were compared between cases and controls using χ² tests. Mean differences between cases and controls were examined by analysis of covariance, allowing for the matching after converting serum levels of Hsp70 and hsCRP to logarithmic values. Conditional logistic regression models with sex, age, and study area strata were applied to calculate odds ratios (OR) and 95% confidence intervals (95% CI) for lung cancer. ORs were computed according to quartile levels of serum Hsp70 and hsCRP. Cases were categorized into four groups according to the quartile in controls for serum Hsp70 and hsCRP. Potential confounding factors considered included smoking habits, drinking habits (never drinker, former drinker, current drinker, or unknown), and BMI (<20.0, 20.0-24.9, ≥25.0 kg/m², or unknown). As smoking represents an extremely important risk factor for lung cancer, adjustments were made in analyses of all and male subjects using the following detailed strata: subjects who had never smoked; former smokers who had not smoked for 0 to 4, 5 to 9, 10 to 14, 15 to 19, or ≥20 years; current smokers with pack-years of 0-19, 20-39, 40-59, 60-79, 80-99, or ≥100; and unknown smoking habits. In analyses of female subjects, adjustments were made using the following smoking strata: never smoker, former smokers, current smokers, and unknown smoking habits. Two-sided Ps < 0.05 were considered statistically significant. All statistical analyses were done using the Statistical Analysis System.

Table 1 shows baseline characteristics for study subjects. For both sexes, current smokers comprised a much greater proportion of cases than controls, and no significant differences between cases and controls were observed for age distribution and drinking habits. In males and females, the proportion of high BMI (≥25.0 kg/m²) was significantly lower in cases than in controls.

Table 2 compares serum levels of Hsp70 and hsCRP between cases and controls. Serum Hsp70 levels were significantly higher in cases than in controls. For males, serum Hsp70 levels tended to be higher in cases than in controls. Serum hsCRP levels did not differ significantly between cases and controls in either sex.

Table 3 shows ORs and 95% CIs for lung cancer according to serum levels of Hsp70 and hsCRP after adjusting for confounding factors. ORs (95% CIs) across quartiles for serum Hsp70 adjusted for sex, age, and study area (OR1) were 0.87 (0.49-1.55), 1.47 (0.86-2.51), and 1.67 (0.91-3.08; P_trend = 0.040). OR (95% CI) for serum Hsp70 adjusted for sex, age, study area, smoking and drinking habits, and BMI (OR2) displayed similar trends. Analysis limited to men produced more pronounced findings than those for all subjects. In males, OR2s (95% CIs) across quartiles for serum Hsp70 were 1.30 (0.59-2.84), 1.74 (0.83-3.67), and 2.49 (1.06-5.85; P_trend = 0.029). In females, a similar trend was obtained but was not significant. No significant findings were identified for OR1 and OR2 and trends for serum hsCRP in either sex. Modified data sets, excluding cases with <2 years of follow-up, were also analyzed. Results of these analyses were consistent with results of analyses without exclusion. OR2s (95% CIs) across quartiles for serum Hsp70 were 0.84 (0.43-1.64), 1.57 (0.82-2.99), and 1.84 (0.88-3.85; P_trend = 0.044) in males.

We have observed an increasing risk of lung cancer across increasing quartiles of serum Hsp70 levels in Japanese males. However, this association was not significant in females. Moreover, no significant association was found between serum hsCRP levels and risk of lung cancer in either sex.

Hsps are typically regarded as intracellular molecules with important chaperone and cytoprotective functions (11) and are present in all groups of organisms (12). Expression of Hsps can be markedly up-regulated by all cells under conditions of stress, such as exposure to inflammatory mediators and oxidative stress. Increased expression of Hsp70 on airway cells is observed in patients with asthma (13). Serum Hsp70 levels reportedly exhibit positive relationships with several inflammation markers, such as WBC count, interleukin-6, and tumor necrosis factor (TNF)-α (14).

Hsp70 is overexpressed in various tumors, such as cancers of the lung (15) and colon (16). To avoid the effects of subclinical cancer, the modified data, excluding subjects with <2 years of follow-up, were analyzed. Results of these analyses were consistent with those analyses without exclusion.

Conversely, serum hsCRP levels were not associated with risk of lung cancer in either sex. Serum CRP levels have recently been positively associated with risk of lung cancer in another U.S. population (17). Plasma CRP concentrations were elevated among individuals who subsequently developed colon cancer in a U.S. population (7). However, serum hsCRP levels are not significantly associated with increased risk for colorectal or lung cancer in the JACC Study (18). This lack of positive association may be due to differences in distribution of serum hsCRP levels between U.S. and Japanese populations. Serum CRP levels are lower in Japanese populations than in U.S. populations (19). The absence of any association between serum hsCRP and lung cancer risk in this study might be due to such differences in distributions of serum CRP levels.

Several studies have supported associations between inflammation and risk of lung cancer (20-22). Previous nonmalignant lung disease, such as asthma and chronic bronchitis, has been suggested as a risk factor for lung cancer (21, 22). Asthma may increase risk of lung cancer by reducing the clearance of inhaled toxins and carcinogens in the bronchoalveolar epithelium and by chronic inflammation of the airway or via the immune system (22). Associations have also been reported between infectious inflammation and lung cancer (20). Chlamydia pneumoniae is a common cause of acute respiratory infection and has been hypothesized to cause lung cancer (20). Mediators of inflammation and oxidative stress caused by infection with C. pneumoniae, such as interleukin-1β, interleukin-8, TNF-α, and reactive oxygen species, play important roles in damage to lung tissue and DNA that can result in carcinogenesis (23). Infection with C. pneumoniae may localize more easily in the lungs of smokers and could also act synergistically with cigarette smoking to increase the risk of lung cancer (24).

Several recent epidemiologic studies have reported associations between inflammation biomarkers and risk of lung cancer (17, 25). Serum CRP is reportedly associated with lung cancer even after adjustment for smoking, but interleukin-6 and tumor necrosis factor-α showed no significant association (25). The association between WBC and risk of lung cancer was also significant and positive before adjusting for smoking and other potential confounders (25). This association was attenuated after adjustment for smoking, but significance for trends remained. In our study, however, associations between Hsp70 and risk of lung cancer were not attenuated after adjustment for smoking habits.

In female subjects in the present study, no association was observed between Hsp70 or hsCRP and lung cancer risk. It has been well known that sex hormones play a role in the inflammation. Inflammation response might be differences between males and females. However, the analysis in only female subjects lacked adequate statistical power. These associations should be examined again in a larger sample, as the number of female lung cancer cases was insufficient in this study.

In conclusion, the present results show that increased levels of serum Hsp70 represent a risk factor for lung cancer among Japanese. Although further studies are needed to clarify associations between chronic inflammation and lung cancer, serum Hsp70 levels may offer a more attractive biomarker for predicting risk of lung cancer than serum hsCRP levels.

The present members of the JACC Study and their affiliations are as follows (9): Dr. Akiko Tamakoshi (Present Chairman of the JACC Study Group), Nagoya University Graduate School of Medicine; Dr. Mitsuru Mori, Sapporo Medical University School of Medicine; Dr. Yutaka Motohashi, Akita University School of Medicine; Dr. Ichiro Tsuji, Tohoku University Graduate School of Medicine; Dr. Yosikazu Nakamura, Jichi Medical School; Dr. Hiroyasu Iso, Osaka University Graduate School of Medicine; Dr. Haruo Mikami, Chiba Cancer Center; Dr. Yutaka Inaba, Juntendo University School of Medicine; Dr. Yoshiharu Hoshiyama, University of Human Arts and Sciences Graduate School; Dr. Hiroshi Suzuki, Niigata University Graduate School of Medical and Dental Sciences; Dr. Hideaki Toyoshima, Nagoya University Graduate School of Medicine; Dr. Shinkan Tokudome, Nagoya City University Graduate School of Medicine; Dr. Yoshinori Ito, Fujita Health University School of Health Sciences; Dr. Shuji Hashimoto, Fujita Health University School of Medicine; Dr. Shogo Kikuchi, Aichi Medical University School of Medicine; Dr. Kenji Wakai, Aichi Cancer Center Research Institute; Dr. Akio Koizumi, Graduate School of Medicine and Faculty of Medicine, Kyoto University; Dr. Takashi Kawamura, Kyoto University Center for Student Health; Drs. Yoshiyuki Watanabe and Tsuneharu Miki, Kyoto Prefectural University of Medicine Graduate School of Medical Science; Dr. Chigusa Date, Nara Women's University Graduate School of Humanities and Sciences; Dr. Kiyomi Sakata, Wakayama Medical University; Dr. Takayuki Nose, Tottori University Faculty of Medicine; Dr. Norihiko Hayakawa, Research Institute for Radiation Biology and Medicine, Hiroshima University; Dr. Takesumi Yoshimura, Fukuoka Institute of Health and Environmental Sciences; Dr. Akira Shibata, Kurume University School of Medicine; Dr. Naoyuki Okamoto, Kanagawa Cancer Center; Dr. Hideo Shio, Moriyama Municipal Hospital; Dr. Yoshiyuki Ohno (Former Chairman of the JACC Study Group), Asahi Rosai Hospital; Dr. Tomoyuki Kitagawa, Cancer Institute of the Japanese Foundation for Cancer Research; Dr. Toshio Kuroki, Gifu University; and Dr. Kazuo Tajima, Aichi Cancer Center Research Institute.

The previous investigators of the study group are listed in ref. 8, except for the following nine members (affiliations are those at the time of participation in the study): Dr. Takashi Shimamoto, Institute of Community Medicine, University of Tsukuba; Dr. Heizo Tanaka, Medical Research Institute, Tokyo Medical and Dental University; Dr. Shigeru Hisamichi, Tohoku University Graduate School of Medicine; Dr. Masahiro Nakao, Kyoto Prefectural University of Medicine; Dr. Takaichiro Suzuki, Research Institute, Osaka Medical Center for Cancer and Cardiovascular Diseases; Dr. Tsutomu Hashimoto, Wakayama Medical University; Dr. Teruo Ishibashi, Asama General Hospital; Dr. Katsuhiro Fukuda, Kurume University School of Medicine; and Dr. Hiroyuki Shimizu, Gifu University School of Medicine.

We thank Dr. Kunio Aoki (Professor Emeritus of Nagoya University School of Medicine and Former Chairman of the JACC Study Group) and Dr. Haruo Sugano (Former Director of the Cancer Institute of the Japanese Foundation for Cancer Research) who greatly contributed to the initiation of the study.