Relationship between Arthritis and Leukemia Not Causal
By FD GROVES AND RF VOGT •
April 21, 2000
If there is an increased risk of subsequent CLL among patients with rheumatoid arthritis and other autoimmune disorders, the association may not necessarily be causal. It is possible that certain individuals are genetically predisposed to develop one or both of these disorders. The authors concluded that “These data...support the hypothesis that genetic factors disturbing the regulation of the immune system may predispose both to lymphoid neoplasms and to autoimmune disease.” Thus, it seems reasonable to speculate that the two diseases may be related to each other,…
However, the epidemiologic studies to date have been contradictory, with recent negative studies casting doubt on the early reports of a causal association. Further study of B-cell monoclonal lymphocytosis may help to shed light on the relationship, if any, between autoimmune and lymphoproliferative disorders.
The possibility of a causal relationship between autoimmune disorders and subsequent leukemia was first entertained forty years ago. British cohort studies (1,2) in the late 1950s revealed an increased risk of leukemia among patients irradiated for ankylosing spondylitis (rheumatoid arthritis of the spine). This led investigators to question whether it was the radiation treatment which increased the risk of subsequent leukemia, or whether autoimmune disease per se predisposed to leukemia. A case-control study (3) implicated both “rheumatic diseases” and irradiation as independent risk factors for leukemia in British servicemen (N=679 cases and 1367 controls) diagnosed between 1939 and 1955. A more comprehensive case-control study (4) later found that all of the rheumatic diseases increased the risk of all “reticuloses”, including leukemia, “lymphosarcoma”, and Hodgkin's disease, diagnosed in British servicemen (N=1356 cases and 1378 controls) between 1939 and 1958.
Subsequent studies initially supported the hypothesis of a causal relationship between autoimmune disorders and lymphoproliferative (relating to the production of lymphoid tissue) disorders. For example, a nationwide population-based linked-registry cohort study of rheumatoid arthritis patients (N=11,483 males and 34,618 females) in Finland (5,6) revealed increased risks of Hodgkin's disease (females only, 14 cases), non-Hodgkin's lymphomas (both males, 13 cases, and females, 25 cases), myeloma (both males, 7 cases, and females, 21 cases), and leukemia (males only, 18 cases). A smaller (N=489) clinic-based cohort study of British rheumatoid arthritis patients (7) found an increased risk of dying from leukemia (3 cases) or “lymphosarcoma” (5 cases).
These early studies did not specify the leukemia cell types (eg, CLL- chronic lymphocytic leukemia) for which rheumatoid arthritis patients were alleged to be at increased risk. However, more specific studies soon filled the void, linking various pre-existing diseases, both autoimmune and otherwise, with increased risk of subsequent CLL. For example, a population-based case-control study (8) of white adult
leukemia patients (N=450 CLL cases and 1237 random controls) from the Tri-State Leukemia Survey (in upstate New York, Baltimore, and Minneapolis/St. Paul between 1959 and 1962) revealed increased risks of subsequent CLL among male patients with pre-existing asthma, “hay fever”, hives, eczema, diabetes mellitus, tuberculosis, pneumonia, heart disease, or “rheumatism”, and among patients of both sexes with pre-existing “goiter”, herpes zoster, psoriasis, or neurodermatitis. A substudy (9) focusing on just the males (N=234 CLL cases and 668 random controls) from the Tri-State Survey confirmed a statistically-significant increased risk of CLL among those with a history of “rheumatism and arthritis” (28 cases) or eczema (11 cases). A later hospital-based matched case-control study (N=342 cases and 342 controls) in Baltimore
(10) between 1975 and 1982, however, found no increased risk of CLL in patients with a variety of pre-existing chronic infectious or autoimmune diseases. On the contrary, two protective factors were identified: history of allergies and prior surgical excision of lymphoid tissue (eg, tonsillectomy, adenoidectomy, or appendectomy).
Three successively larger hospital-based matched case-control studies of lymphoid malignancies in the Yorkshire region of England likewise found no evidence for an association between autoimmune diseases and CLL. A pilot study (11) in Yorkshire between 1979 and 1981 found non-significantly increased risks of CLL in patients (N=66 CLL cases and 66 controls) with prior history of tuberculosis or infectious mononucleosis; neither tonsillectomy nor appendectomy appeared to be protective. A follow-up study (12) in Yorkshire between 1979 and 1984 found that CLL risk was increased in patients (N=245 CLL cases and 417 controls) of both sexes with a past medical history of cancer (24 cases), scarlet fever (11 cases), herpes zoster (36 cases), bronchitis (21 cases), and chronic ear infection (23 cases); in male patients with a past medical history of hypertension (37 cases) or myocardial infarction (15 cases); and in female patients with a past medical history of migraine (10 cases) or osteoarthritis; rheumatoid arthritis was not associated with CLL, and appendectomy (52 cases) appeared to be protective. The third and final study (13) in Yorkshire between 1979 and 1986, found an increased risk of CLL in patients (N=245 CLL cases and 417 controls) with a family history of multiple sclerosis; past medical history of cancer, radiotherapy, heart disease, heart drugs, or skin lesions; and occupational exposure to live animals. By the time this paper was published in 1990, autoimmunity was not even mentioned as a risk factor.
Similarly, a population-based case-control study (14) of CLL (N=430 cases and 1681 matched controls) in four US metropolitan areas (Atlanta, Detroit, Salt Lake City, and Seattle) between 1977 and 1981 found “little evidence of a relation between chronic antigenic stimulation and the occurrence of chronic lymphocytic leukemia”, although the risk of CLL was increased among patients with a past history of fever blisters (223 cases), chronic urinary tract infections (125 cases), syphilis (9 cases), tuberculosis (16 cases), and rheumatic fever (10 cases). By the time this paper was published in 1991, autoimmune disorders were not even considered as CLL risk factors in the course of the study. A small (N=98 CLL cases and 123 controls) case-control study (15) of hematopoietic (relating to the production of blood in the body) malignancies among members of the Kaiser/Permanente Medical Care Program in Northern California and the Pacific Northwest found an increased risk of CLL among patients with a history of rheumatic fever (4 cases). However, the authors concluded that “...[T]his study provided little, if any, support for an association of chronic infectious, autoimmune, allergic, and musculoskeletal conditions with subsequent occurrence of the leukemias or [non-Hodgkin's lymphoma]”, and that “...[T]hese data did not support a role for chronic antigenic stimulation, as defined in previous epidemiologic studies, in the etiology of hematopoietic malignancies.” A linked registry cohort study (16) of rheumatoid arthritis patients (N=11,683) hospitalized in Sweden between 1965 and 1983 found an increased risk of CLL in men (nine cases), but not in women (two cases); when data for both sexes were combined, the relative risk was not significantly increased.
All of the foregoing studies have been conducted in Caucasian populations. However, a population-based case-control study (17) of adult leukemias in Shanghai, China between 1987 and 1989 found increased risks of CLL, albeit based on small numbers (N=21 CLL cases and matched controls) among patients with a prior history of rheumatoid arthritis (two cases), hyperthyroidism (one case), chronic infections (five cases) other than tuberculosis (this category included chronic bronchitis, sinusitis, cholecystitis, and pyelonephritis), appendectomy (five cases), and use of salicylates (two cases).
If there is an increased risk of subsequent CLL among patients with rheumatoid arthritis and other autoimmune disorders, the association may not necessarily be causal. It is possible that certain individuals are genetically predisposed to develop one or both of these disorders. This hypothesis was tested in a case-control study of the family histories of 28 CLL patients and 28 unmatched controls from the Johns Hopkins Hospital (18). Among 320 relatives of the 28 CLL cases, seven had hematopoietic malignancies including four with CLL, two with unspecified leukemias, and one with “lymphosarcoma”. Eighteen relatives of CLL cases had various autoimmune disorders including six with hyperthyroidism, four with rheumatoid arthritis, three with systemic lupus erythematosus, and two with pernicious anemia. Among 396 relatives of the 28 hospital controls, one had CLL and four had autoimmune diseases, including two with pernicious anemia, one with hyperthyroidism, and one with rheumatoid arthritis. The authors concluded that “These data...support the hypothesis that genetic factors disturbing the regulation of the immune system may predispose both to lymphoid neoplasms and to autoimmune disease.”
The lymphocytes of some CLL patients have surface immunoglobulin M (IgM) with rheumatoid factor (anti-IgG) activity (19, 20, 21). Conversely, some rheumatoid arthritis patients have circulating CD5+ B-cells resembling those found in CLL (22). Thus, it seems reasonable to speculate that the two diseases may be related to each other, or at least that they may be related to the B-cell monoclonal lymphocytosis described elsewhere in this volume. However, the epidemiologic studies to date have been contradictory, with recent negative studies casting doubt on the early reports of a causal association. Further study of B-cell monoclonal lymphocytosis may help to shed light on the relationship, if any, between autoimmune and lymphoproliferative disorders.
1. Abbatt JD and Lea AJ: Lancet 1956; 2:1317.
2. Court-Brown WM and Doll R: Leukemia and aplastic anemia in patients irradiated for
ankylosing spondylitis. Medical Research Council, Special Report Series No. 295; London,
HM Stat Office, 1957.
3. Abbatt JD and Lea AJ: Leukaemogens. Lancet 1958; 2:880-883.
4. Lea AJ: An association between the rheumatic diseases and the reticuloses. Ann Rheum Dis
5. Isomaki HA, Hakulinen T, and Joutsenlahti U: Excess risk of lymphomas, leukemia, and
myeloma in patients with rheumatoid arthritis. J Chron Dis 1978; 31:691-696.
6. Hakulinen T, Isomaki H, and Knekt P: Rheumatoid arthritis and cancer studies based on
linking nationwide registries in Finland. Am J Med 1985; 78(Suppl 1A):29-32.
7. Prior P, Symmons DPM, Hawkins CF, et al.: Cause of death in rheumatoid arthritis. Br J
Rheumatol 1984; 23:92-99.
8. Viadana E and Bross ID: Use of the medical history to predict the future occurrence of
leukemias in adults. Prev Med 1974; 3:165-170.
9. Gibson R, Graham S, Lilienfeld A, Schuman L, Levin M, and Swanson M: Epidemiology of
diseases in adult males with leukemia. J Natl Cancer Inst 1976; 56:891-898.
10. Linet MS, McCaffrey LD, Humphrey RL, Brookmeyer R, VanNatta ML, Tielsch JM, Bias
WB, Markowitz JA, Kravitz SC, and Szklo M: Chronic lymphocytic leukemia and acquired
disorders affecting the immune system: A case-control study. J Natl Cancer Inst 1986;
11. Bernard SM, Cartwright RA, Bird CC, Richards IDG, Lauder I, and Roberts BE: Aetiologic
factors in lymphoid malignancy: A case-control epidemiological study. Leuk Res 1984;
12. Cartwright RA, Bernard SM, Bird CC, Darwin CM, O'Brien C, Richards IDG, Roberts B,
and McKinney PA: Chronic lymphocytic leukemia: Case-control epidemiological study in
Yorkshire. Br J Cancer 1987; 56:79-82.
13. McKinney PA, Alexander FE, Roberts BE, O'Brien C, Bird CC, and Cartwright RA:
Yorkshire case-control study of leukemias and lymphomas: Parallel nationwide analysis of
seven disease categories. Leuk Lymph 1990; 2:67-80.
14. Rosenblatt KA, Koepsell TD, Daling JR, Lyon JL, Swanson GM, Greenberg RS, and Weiss
NS: Antigenic stimulation and the occurrence of chronic lymphocytic leukemia. Am J
Epidemiol 1991; 134:22-28.
15. Doody MM, Linet MS, Glass AG, Friedman GD, Pottern LM, Boice JD, and Fraumeni JF:
Leukemia, lymphoma, and multiple myeloma following selected medical conditions. Cancer
Causes Control 1992; 3:449-456.
16. Gridley G, McLaughlin JK, Ekbom A, Klareskog L, Adami HO, Hacker DG, Hoover R,
and Fraumeni JF: Incidence of cancer among patients with rheumatoid arthritis. J Natl Cancer
Inst 1993; 85(4):307-311.
17. Zheng W, Linet MS, Shu XO, Pan RP, Gao YT, and Fraumeni JF: Prior medical conditions
and the risk of adult leukemia in Shanghai, People's Republic of China. Cancer Causes
Control 1993; 4:361-368.
18. Conley CL, Misiti J, and Laster AJ: Genetic factors predisposing to chronic lymphocytic
leukemia and to autoimmune disease. Medicine 1980; 59(5):323-334.
19. Preud'homme JL and Seligmann M: Anti-human Immunoglobulin G activity of
membrane-bound monoclonal immunoglobulin M in lymphoproliferative disorders. Proc Natl
Acad Sci 1972; 69(8):2132-2135.
20. Borche L, Lim A, Binet JL, and Dighiero G: Evidence that chronic lymphocytic leukemia
B-lymphocytes are frequently committed to production of natural autoantibodies. Blood
21. Katsakis PD, Pavlidis NA, Papadopoulos NM, Bai M, and Moutsopoulos HM: Monoclonal
Immunoglobulins and autoantibodies in lymphoid malignancy. Anticancer Research 1990;
22. Plater-Zyberk C, Maini RN, Lam K, Kennedy TD, Janossy G: A rheumatoid arthritis B-cell
subset expresses a phenotype similar to that in chronic lymphocytic leukemia. Arthritis and
Source: The Centers for Disease Control and Prevention
Merriam Webster’s Medical Desk Dictionary, Springfield: Merriam Webster Inc., 1993.