Drs. Chen and Chang contributed equally to this work.
The risk of cancer in patients with rheumatoid arthritis: A nationwide cohort study in Taiwan†
Article first published online: 28 JAN 2011
Copyright © 2011 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 63, Issue 2, pages 352–358, February 2011
How to Cite
Chen, Y.-J., Chang, Y.-T., Wang, C.-B. and Wu, C.-Y. (2011), The risk of cancer in patients with rheumatoid arthritis: A nationwide cohort study in Taiwan. Arthritis & Rheumatism, 63: 352–358. doi: 10.1002/art.30134
This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health and managed by the National Health Research Institute of Taiwan. The interpretations and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institute.
- Issue published online: 28 JAN 2011
- Article first published online: 28 JAN 2011
- Accepted manuscript online: 4 NOV 2010 01:01PM EST
- Manuscript Accepted: 28 OCT 2010
- Manuscript Received: 29 JUN 2010
- Taichung Veterans General Hospital
The association of rheumatoid arthritis (RA) and malignancy has rarely been explored in Asian populations. The aim of this study was to investigate the relative risk of cancer in Taiwanese patients with RA and to identify groups of patients with a high risk of cancer.
We conducted a nationwide cohort study of the risk of cancer among 23,644 patients with RA who had no history of malignancies, using the National Health Insurance database of Taiwan from 1996 to 2007. Standardized incidence ratios (SIRs) for various cancers were analyzed.
Among the patients with RA, 935 cancers were observed. Patients with RA had an increased risk of cancer (SIR 1.23, 95% confidence interval [95% CI] 1.22–1.23), especially hematologic cancers (SIR 2.74, 95% CI 2.68–2.81). The relative risk of cancer was higher among younger patients. Most cancer cases were detected within the first year following the diagnosis of RA. The relative risk of cancer decreased as the duration of observation increased. Among hematologic cancers, the risk of non-Hodgkin's lymphoma was greatest (SIR 3.54, 95% CI 3.45–3.63). Among solid tumors, the risk of cancers of the kidney and vagina/vulva was highest. A decreased risk of cancers of the cervix and nonmelanoma skin cancer in patients with RA was also observed.
Patients with RA have an increased risk of cancer, especially hematologic and kidney cancers. The relative risk of cancer in patients with RA decreased with long-term followup. Cancer screening with continued vigilance is recommended for patients with RA.
Rheumatoid arthritis (RA) affects several organ systems, including the cardiovascular, musculoskeletal, and respiratory systems, and is associated with some psychological problems, all of which leads to major comorbidities and mortality. Chronic organ damage and late complications such as malignancy and cardiovascular disease are major adverse factors associated with both morbidity and mortality (1). However, the association of malignancies with RA remains controversial. Some studies have demonstrated a similar incidence of malignant diseases in patients with RA and individuals in the general population (2–6), yet other reports have indicated a lower risk of malignant diseases in patients with RA (7, 8). Most studies have consistently indicated an increased incidence of lymphoma (3, 5, 9, 10) and a decreased incidence of colorectal cancer in patients with RA (2, 7, 9). Because of the relatively low prevalence of RA in Asian populations (11–13), the relationship between malignancy and RA has rarely been reported in these countries (14).
The pathogenetic mechanisms of interaction between RA and cancer remain uncertain. Baecklund et al (15) first demonstrated that chronic inflammation, rather than its treatment, is a major determinant of lymphoma risk in patients with RA. However, other factors such as treatment modalities may also be involved (16). The aim of this study was to investigate the relative risk of malignant diseases, including specific cancer types, after the diagnosis of RA in Taiwanese patients, using a nationwide cohort database.
PATIENTS AND METHODS
This study was based on data from the National Health Insurance Research Database (NHIRD) released by the National Health Research Institute. Taiwan began its National Health Insurance program in 1995 to finance health care for all of its residents. More than 25 million individuals are enrolled in the program, representing ∼99% of the entire population of Taiwan. The database comprises comprehensive information on insured persons, such as demographic data, dates of clinical visits, diagnostic codes, details of prescriptions, and expenditure amounts, as described in detail in our previous studies (17–21). During the study period, International Classification of Diseases, Ninth Revision (ICD-9) codes were used to define diseases. Personal information including family history, lifestyle, and habits such as smoking and alcohol use was not available from the NHIRD.
Patients with RA.
All cases of RA were obtained from the Registry of Catastrophic Illness Database, a subpart of the NHIRD. In Taiwan, insured persons with major diseases can apply for a catastrophic illness certificate that grants exemption from co-payment. The issuance of catastrophic illness certificates was validated by at least 2 specialists, based on careful examination of the medical records, laboratory studies, and imaging studies. Only individuals who meet the diagnostic criteria for major diseases are issued a catastrophic illness certificate. RA and cancer are statutorily included in the catastrophic illness category. Both the outpatient and inpatient claims of beneficiaries with a catastrophic illness certificate are collected in the catastrophic illness profile and distributed as a package. The prescription claims of beneficiaries are released in a different data set and are not included in current catastrophic illness profiles.
Patients with RA (ICD-9 code 714.0) were included between January 1, 1996 and December 31, 2005. Patients younger than age 16 years were excluded. All of the included patients were followed up until December 31, 2007 to ensure that each included patient had a sufficient observation time (at least 2 years). Application of a catastrophic illness certificate for RA requires thorough clinical and laboratory evaluations, the results of which fulfill the American College of Rheumatology criteria for the classification of RA (22). Patients with other autoimmune diseases such as systemic lupus erythematosus or Sjögren's syndrome were excluded. Patients with a history of malignancies were also excluded.
Because the data set used in this study consists of de-identified secondary data released to the public for research purposes, this study was exempt from full review by the Institutional Review Board.
Identification of cancer cases.
We identified the diagnoses of cancers using records from the Registry of Catastrophic Illness Database. To apply for a cancer catastrophic illness certificate, patients should provide cytology or pathology reports or evidence such as additional laboratory and imaging studies supporting the diagnosis of cancer, including tumor marker surveys, radiographs, bone scans, computed tomography scans, or magnetic resonance imaging scans. At least 2 other oncologists carefully examine the medical records and laboratory information, including imaging studies. Only patients who meet the criteria for the diagnoses are issued certificates. We excluded patients with in situ malignancies, because patients with in situ malignant diseases do not qualify for a catastrophic illness certificate. The diagnostic codes for malignancies were defined as codes from 140 to 208.91 in the ICD-9 clinical modification format. We categorized these cancer cases into hematologic cancers and nonhematologic cancers. Hematologic cancers were subcategorized into leukemias (coded 204-208) and lymphomas (including non-Hodgkin's lymphoma [coded 200, 202-203] and Hodgkin's lymphoma [coded 201]), according to the methods used by the Cancer Registry in Taiwan.
Cancer risk and statistical analysis.
All study subjects were followed up until a first-time diagnosis of cancer, death, the end of followup in the medical records, or the end of 2007. We examined the association between RA and specific cancer types using standardized incidence ratios (SIRs). The SIR was calculated as follows: the number of cancer cases among patients with RA divided by the expected number of cancer cases according to national age-specific, sex-specific, and period-specific cancer rates. Yearly reports of cancer rates were obtained from the Taiwan National Cancer Registry. We pooled the 10-year Cancer Registry reports for Taiwan from 1996 to 2006 as a standard.
To assess the effect of age on the relative risk of malignancies, we analyzed the relative risk among patients who were ages 0–39 years, 40–69 years, and ≥70 years at the time of the diagnosis of RA. A further analysis was done to evaluate whether the association of malignancies varied according to the length of followup after RA was diagnosed. We divided followup time into the following 6 periods: ≤1 year, 1–2 years, 2–4 years, 4–6 years, 6–8 years, and >8 years.
SAS version 9.1 was used to perform the statistical analysis of the data in this study.
Risk of cancer in patients with RA who have no history of malignancy.
We identified a total of 23,644 patients with RA who did not have previous malignancies. The mean ± SD age of the patients at the time of the diagnosis of RA was 53.08 ± 14.38 years. The RA study group comprised 18,527 female patients (78.36%) and 5,117 male patients (21.64%). The mean ± SD followup time for the patients with RA was 5.90 ± 2.87 years (Table 1). During the observation period of 139,555.47 person-years, a total of 935 cancers were identified after the diagnosis of RA. A slightly increased overall cancer risk in patients with RA was observed (SIR 1.23, 95% confidence interval [95% CI] 1.22–1.23) (Table 2). The risk of developing cancer was similar in women and men.
|Age at diagnosis, mean ± SD||53.08 ± 14.38|
|Duration of followup, mean ± SD years||5.90 ± 2.87|
|Characteristic||All cancers†||Hematologic cancers|
|No. observed||No. expected||SIR||95% CI||No. observed||No. expected||SIR||95% CI|
The relative risk of cancer was greatest among younger patients with RA, and the risk ratio declined with age (Table 2). The SIRs for cancer were 2.35 (95% CI 2.28–2.42) among patients younger than age 40 years, 1.58 (95% CI 1.56–1.59) among those between the ages of 40 years and 69 years, and 1.03 (95% CI 1.01–1.04) among patients older than age 70 years.
Malignant diseases in patients with RA were mostly detected during the first year following the diagnosis of RA (SIR 58.96, 95% CI 58.13–59.96). The risk ratios gradually reduced as the duration of observation increased (Table 2). After 8 years of followup, the SIR for malignancies in patients with RA fell to 0.31, which is less than that in the general population.
Specific types of nonhematologic cancers associated with RA.
A slightly increased risk of nonhematologic malignancies (SIR 1.12, 95% CI 1.11–1.13) was also observed (Table 3). The highest risks were associated with cancers originating from the kidney and female genitalia, such as the vulva and vagina (Table 3). Other significantly associated cancers in RA included melanoma and those originating from the nasopharynx, the thyroid gland, and the lungs (Table 3). In contrast, a reduced risk of certain cancers was also observed, including cancers of the colorectum and cervix and nonmelanoma skin cancers (Table 3).
|Specific cancer type||No. observed||No. expected||SIR||95% CI|
|Nasopharynx, sinus, ears||22||15.79||1.39||1.34–1.45|
|Lung and mediastinum||123||90.47||1.36||1.34–1.38|
|Liver and gallbladder||126||108.59||1.16||1.14–1.18|
|Oropharynx and larynx||31||29.26||1.06||1.02–1.10|
|Cancer of ill-defined sites||14||13.79||1.02||0.96–1.07|
|Colon and rectum||102||107.99||0.94||0.93–0.96|
Hematologic cancers associated with RA.
Among all comorbid malignant diseases, there were 75 cases (8.02%) of hematologic malignancies, including 59 cases of non-Hodgkin's lymphoma, 1 case of Hodgkin's lymphoma, and 15 cases of leukemia. A significantly elevated risk of hematologic malignancies in RA was observed (SIR 2.74, 95% CI 2.68–2.81), especially in male patients (Table 2). The risk of cancer varied by age. Middle-aged patients (ages 40–69 years) had the greatest relative risk of hematologic cancers compared with persons in the general population with similar ages (Table 2).
The risks of hematologic cancers were extraordinarily high during the first 2 years after the diagnosis of RA. The relative risk of hematologic cancer declined as the duration of observation increased (Table 2). The SIR declined to 0.62 after 8 years of followup. With regard to specific cancer types, the risk of non-Hodgkin's lymphoma (including lymphosarcoma, reticulosarcoma, multiple myeloma, and other immunoproliferative neoplasms) was greatest (SIR 3.54, 95% CI 3.45–3.63), followed by Hodgkin's lymphoma (SIR 1.76, 95% CI 1.45–2.17) and leukemia (SIR 1.48, 95% CI 1.41–1.56) (Table 4).
|Cancer type||No. observed||No. expected||SIR||95% CI|
|Non-Hodgkin's lymphoma and others†||59||16.66||3.54||3.45–3.63|
This is the first large-scale nationwide study to estimate the risk of cancer in a population of Taiwanese patients with RA who had no history of malignancies. Consistent with prior cohort studies, the risks of hematologic cancers including Hodgkin's lymphoma, non-Hodgkin's lymphoma, and leukemia were observed to be significantly higher in patients with RA compared with the general population. In addition to being at risk of hematologic cancers, patients with RA had a greater risk for the development of some less common cancers such as cancers of the kidney, vulva, thyroid gland, and nasopharynx compared with the general population.
Most cancer cases in our cohort were detected within the first 2 years after the diagnosis of RA. Aggressive surveillance for cancer during that period of time may have resulted in detection bias. However, when cancer cases observed in the first 2 years were excluded, there was a sustained elevated risk among patients with RA until 8 years of followup, indicating a true link between these 2 diseases. Although the association between malignancy and RA remains controversial, the association of lymphoma with RA has been frequently observed. In addition to common genetic factors between autoimmune diseases and lymphomas, aberrant Epstein-Barr virus expression, persistent inflammation, chronic activation of autoimmune B cells, and immunosuppressive therapy have been proposed to contribute to the development of lymphoma (23, 24). The assessment of lymphoma risk in association with immunomodulating therapy in patients with autoimmune disease is complex due to the inability to distinguish therapy-related effects from effects of the disease itself. The relationship between methotrexate (15, 25–28), nonsteroidal antiinflammatory drugs (NSAIDs), aspirin (29, 30), or corticosteroids (16, 31, 32) and lymphoma risk has been explored in patients with RA, but the results have been inconsistent. We were not able to examine this association, because we did not have information regarding the treatment of the patients with RA in our study.
The finding of higher cancer risks in younger patients with RA was not reported until recently (33). Because the incidence of malignant diseases increases with age in the general population, it is reasonable that the higher prevalence of cancer among older patients makes the difference in incidence between elderly patients with RA and elderly persons in the general population nonsignificant.
Most of the data from previous cohort studies have shown an average 2-fold risk of lymphoma in patients with RA compared with that in the general population (3, 34–36). Our results demonstrated risks similar to those observed in previous studies of non-Hodgkin's lymphoma (SIR 3.54), Hodgkin's lymphoma (SIR 1.76), and, to a lesser extent, leukemia (SIR 1.48) among patients with RA. We did not subcategorize our patients with non-Hodgkin's lymphoma by subtype, because standardized cancer estimates according to specific lymphoma subtypes were not available.
Increased risks of various solid tumors were observed in our patients with RA. In addition to an increased risk of lung cancer (SIR 1.36), significantly elevated risks for certain less-common cancers were observed, such as cancer of the kidney (SIR 2.12), melanoma (SIR 1.47), cancer of the thyroid gland (SIR 1.41), and cancer of the nasopharynx (SIR 1.39). The findings of increased risks of these cancers highlight the power of the present study to detect differences in the incidence of rare cancers. Cigarette smoking, which possibly contributes to RA via the production of antibodies recognizing citrullinated proteins (37), is also a strong risk factor for cancers of the lungs, kidney (38), and urinary bladder (39). Other factors linking lung cancer and RA have been proposed, including chronic inflammation (40) and the presence of interstitial lung disease (41, 42). It was not until recently that an association between RA and kidney cancers was demonstrated. A cohort study in a southern European population showed that RA patients with lung or kidney cancer have higher mortality than expected (2). An elevated risk of vagina/vulva cancer (SIR 1.69) was observed in our patients with RA, which may be explained by an increased prevalence of high-risk human papillomavirus (HPV) infection in patients with RA (43). Cytotoxic drugs and biologic agents such as tumor necrosis factor α blockers may also increase the susceptibility to HPV infection in patients with RA (44).
An increased incidence of melanoma has been reported in patients with RA treated with methotrexate (45) and biologic agents (6). Our study also demonstrated an increased risk of melanoma in patients with RA. However, because only 3 cases of melanoma were observed in our RA cohort, a longer observation time or additional pharmacology-related information may be of benefit in exploring the link between these 2 diseases.
The present study has several limitations. First, we did not have personal information for the patients, such as lifestyle, body mass index, smoking habits and alcohol use, or family history of malignancy or other autoimmune diseases, nor did we have information regarding systemic treatment that may contribute to the cancer risk. Second, misclassification of diseases may occur when using an administrative database. To minimize this bias, we enrolled only patients from the catastrophic illness profile package who met the criteria for RA. The number of patients with RA may therefore be underestimated. Finally, most patients with RA undergo regular physical and laboratory evaluations during observation. A surveillance bias may have contributed to some of the increased frequency of cancer in these patients.
Nonetheless, our study provides important information. This is the first large-scale nationwide cohort study of cancer and RA conducted in an Asian population. Patients with RA, especially younger patients, are at risk for the development of certain cancers when compared with the general population. In addition to a high prevalence of hematologic cancers, especially non-Hodgkin's lymphoma, an increased risk of several less-common cancers, such as cancer of the vagina/vulva, kidney, nasopharynx, and prostate, and melanoma was also observed. Finally, the relative risk of cancer in patients with RA gradually declined as the observation time increased. A decrease in disease activity, long-term exposure to immunomodulatory drugs, or early mortality among patients with severe diseases or disease complicated by comorbidities may contribute to the trend toward a decreased relative risk of cancer over time.
Current treatments for RA in Taiwan are a combination of several immunomodulatory drugs, mainly prednisolone, various NSAIDs, methotrexate, and biologic agents (adalimumab or etanercept). Other disease-modifying antirheumatic drugs such as azathioprine, hydroxychloroquine, cyclophosphamide, and mycophenolate mofetil are occasionally used. Because some of these medications may result in end-organ damage or cancer risks, the direct effect of immunosuppressive treatment on cancer risk in patients with RA needs further studies.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Wu had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Chen, Chang, Wu.
Acquisition of data. Chen, Wang.
Analysis and interpretation of data. Chen, Wang, Wu.
- 43Prevalence and cervical human papilloma virus associated factors in patients with rheumatoid arthritis. Ginecol Obstet Mex 2008; 76: 9–17. In Spanish., , , , , , et al.