Abstract
Purpose
Although a few previous studies have reported positive associations between adult myeloid leukemia and a history of certain medical conditions, the etiology of most cases remains largely unknown. Our purpose was to examine associations between certain medical conditions and adult myeloid leukemia.
Methods
Using logistic regression, we evaluated associations between 16 self-reported medical conditions and myeloid leukemia in a case–control study of 670 cases [including 420 acute myeloid leukemia (AML) and 186 chronic myelogenous leukemia (CML)] and 701 population-based controls.
Results
We observed significant positive associations between AML and ulcerative colitis (odds ratio (OR) = 3.8; 95 % confidence interval (CI), 1.1–13) and between CML and peptic ulcer (OR = 2.0; 95% CI, 1.1–3.8). A personal cancer history increased both AML (OR = 2.6; 95% CI, 1.7–3.9) and CML (OR = 3.5; 95% CI, 2.0–5.8) risk even after excluding individuals who reported prior radiation and/or chemotherapy treatment.
Conclusion
Certain inflammatory medical conditions and a personal history of cancer, independent from therapy, are associated with an increased risk of myeloid leukemia.
Keywords: Acute myeloid leukemia, Chronic myeloid leukemia, Autoimmune disorders, Medical conditions
Introduction
Positive associations between adult leukemia and medical conditions including inflammatory bowel diseases (ulcerative colitis (UC) and Crohn’s disease), rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), polymyalgia, cardiovascular disease, celiac disease, and pernicious anemia have been reported. However, with the exception of one large US study of autoimmune conditions and risk of adult myeloid leukemia that linked Surveillance, Epidemiology and End Results registry cases to Medicare data [1], most studies have examined cancer incidence in cohorts with medical conditions that have included very few adult myeloid leukemia cases [2–7]. Characterizing myeloid leukemia risks associated with medical conditions may illuminate etiology and assist in identifying high-risk populations for targeted prevention strategies.
Methods
Study population
Cases
Eligibility criteria for cases (see [8] for complete description of study methods) included: (1) alive at contact, (2) Minnesota resident aged 20–79 years at diagnosis during 1 June 2005–30 November 2009, (3) diagnosis of AML, CML, chronic myelomonocytic leukemia (CMML), or other myeloid leukemia, and (4) ability to speak either English or Spanish. The population-based Minnesota state cancer registry identified 1178 pathologically confirmed incident cases of adult myeloid leukemia, of whom 271 were not referred (death soon after diagnosis (N = 230), physician and/or patient refusal (N = 41)). Of the 907 referrals, we were unable to confirm contact information for 52 patients, and 22 died before contact, resulting in contacts with 833 patients, of whom 16 % declined to participate and 3 % did not meet eligibility criteria (i.e., cognitively impaired, language, age or diagnosis ineligible, not a Minnesota resident, or deceased), yielding an overall cooperation rate of 83 % and a participation rate of 58 % [9]. Central review determined final pathologic-genetic classification of all cases, resulting in 420 AML (3 additional AML cases enrolled but died before questionnaire completion), 186 CML, and 64 CMML/other myeloid leukemia (including not otherwise specified) cases. Study results are given for myeloid leukemia overall, AML, and CML.
Controls
Controls were identified through the Minnesota State Driver’s License/Identification Card list containing nearly all (>98 %) Minnesota residents aged 16–85 years (unpublished data by our group). Controls were selected on a quarterly basis throughout the study and were frequency-matched to cases on age in deciles. Control eligibility criteria were as follows: (1) alive at contact, (2) Minnesota resident, aged 20–79 years, (3) English or Spanish speaker, and (4) no prior myeloid leukemia diagnosis. Of 1020/1200 identified controls contacted, 106 did not meet eligibility criteria, 213 declined to participate, and 701 enrolled, for a total participation rate of 64 %.
Data collection
Medical conditions in cases and controls were assessed through self-report. The 30-min self-administered questionnaire included questions (yes, no, not sure; if yes, age at diagnosis) regarding diagnosis of the following: hyperthyroidism; hypothyroidism; peptic ulcer; ankylosing spondylitis; heart disease, angina or heart attack; high blood pressure; non-gestational diabetes mellitus (Type I vs. II were not ascertained); rheumatoid arthritis; osteoarthritis; Crohn’s disease; UC; celiac disease; Sjogren’s disease or sicca syndrome; lupus or SLE; polymyositis, dermatomyositis, or polymyalgia rheumatica; eczema; contact dermatitis; cirrhosis of the liver or liver damage; infectious mononucleosis; chronic fatigue syndrome; and/or epilepsy (convulsions or seizures not related to high fever). Participants were also asked whether they had ever been diagnosed with cancer including cancer type, diagnosis age, and treatment type(s) (surgery, radiation, chemotherapy, other, none, or don’t know).
The Institutional Review Boards of the University of Minnesota, the Mayo Clinic, and the Minnesota Department of Health, and the 19 participating hospitals approved the study. All participants gave informed consent.
Statistical analysis
Unconditional logistic regression (SAS v9.1, 9.2 (Cary, NC)) was used to model adjusted associations between each medical condition and adult myeloid leukemia if at least 3 cases and 3 controls reported the condition. Potential confounders and interactions between variables were also assessed. All models were adjusted for age decile, sex, and body mass index (categorical); p values ≤ 0.05 were considered significant. To limit the possibility that medical conditions became clinically apparent as a consequence of cancer development, we conducted two sets of analyses where individuals were excluded who reported that their medical condition was diagnosed within one and 3 years of their myeloid leukemia diagnosis or date of questionnaire completion for controls. We also conducted analyses that excluded individuals who reported having a previous cancer diagnosis within 5 years of their leukemia diagnosis or questionnaire completion for controls. A small number of individuals were excluded from regression analyses due to missing data (as noted in Tables 2, 3).
Table 2.
Associations between medical conditions and risk of adult myeloid leukemia
| Controls | Myeloid leukemia |
AML |
CML |
|||||
|---|---|---|---|---|---|---|---|---|
| N (%) | N (%) | OR2 (95% CI) | N (%) | OR2 (95% CI) | N (%) | OR2 (95% CI) | ||
| Autoimmune conditions b | ||||||||
| Any autoimmune conditionc | No | 587 (51.5) | 552 (49.5) | 1.0 | 354 (37.6) | 1.0 | 150 (83.3) | 1.0 |
| Yes | 102 (50.5) | 100 (49.5) | 1.1 (0.8–1.6) | 56 (35.4) | 1.0 (0.7–1.5) | 30 (22.7) | 1.3 (0.8–2.2) | |
| Rheumatoid arthritis | No | 639 (94.0) | 614 (95.3) | 1.0 | 395 (96.6) | 1.0 | 164 (93.2) | 1.0 |
| Yes | 41 (6.0) | 30 (4.7) | 0.8 (0.5–1.3) | 14 (3.4) | 0.6 (0.3–1.1) | 12 (6.8) | 1.3 (0.6–2.5) | |
| Hypothyroidism | No | 635 (94.0) | 603 (93) | 1.0 | 384 (94.6) | 1.0 | 180 (98.4) | 1.0 |
| Yes | 41 (6) | 42 (7) | 1.2 (0.8–2.0) | 22 (5.4) | 1.0 (0.6–1.8) | 14 (7.7) | 1.5 (0.8–3.0) | |
| Hyperthyroidism | No | 655 (96.0) | 627 (97) | 1.0 | 392 (96.8) | 1.0 | 169 (92.4) | 1.0 |
| Ulcerative colitis | No | 683 (99.4) | 641 (98.2) | 1.0 | 402 (98) | 1.0 | 182 (98.9) | 1.0 |
| Yes | 4 (0.6) | 12 (1.8) | 3.5 (1.1–11) | 8 (2.0) | 3.8 (1.1–13) | 2 (1.1) | 2.0 (0.4–11) | |
| Other conditions b | ||||||||
| Heart disease | No | 611 (90.5) | 553 (86.8) | 1.0 | 353 (87.8) | 1.0 | 153 (85.5) | 1.0 |
| Yes | 64 (9.5) | 84 (13.2) | 1.3 (0.9–1.9) | 49 (12.2) | 1.2 (0.8–1.8) | 26 (14.5) | 1.7 (1.0–2.8)d | |
| High blood pressure | No | 418 (63.4) | 397 (62.7) | 1.0 | 258 (64.7) | 1.0 | 104 (59.8) | 1.0 |
| Yes | 241 (37) | 236 (37) | 0.9 (0.7–1.2) | 141 (35) | 0.8 (0.6–1.1) | 70 (40.2) | 1.2 (0.8–1.8) | |
| Diabetes mellitus | No | 611 (52.3) | 557 (87.3) | 1.0 | 352 (87.6) | 1.0 | 153 (86.4) | 1.0 |
| Yes | 71 (10.4) | 81 (12.7) | 1.1 (0.7–1.8) | 50 (12.4) | 1.1 (0.7–1.6) | 24 (13.6) | 1.2 (0.7–2.0) | |
| Osteoarthritis | No | 584 (87.3) | 573 (89.7) | 1.0 | 362 (90.5) | 1.0 | 162 (90) | 1.0 |
| Yes | 85 (12.7) | 66 (10.3) | 0.8 (0.5–1.1) | 38 (9.5) | 0.7 (0.5–1.1) | 18 (10.0) | 0.8 (0.5–1.4) | |
| Eczema | No | 620 (91.5) | 598 (93.2) | 1.0 | 380 (93.8) | 1.0 | 164 (91.6) | 1.0 |
| Yes | 58 (8.5) | 44 (6.9) | 0.8 (0.6–1.3) | 25 (6.2) | 0.8 (0.5–1.2) | 15 (8.4) | 1.1 (0.6–1.9) | |
| Contact dermatitis | No | 650 (95.3) | 626 (96.8) | 1.0 | 398 (97.1) | 1.0 | 171 (96.1) | 1.0 |
| Yes | 32 (4.7) | 21 (3.3) | 0.7 (0.4–1.2) | 12 (2.9) | 0.6 (0.3–1.3) | 7 (3.9) | 0.8 (0.4–1.9) | |
| Peptic ulcer | No | 652 (95) | 608 (94) | 1.0 | 392 (95.8) | 1.0 | 166 (91.2) | 1.0 |
| Yes | 31 (5) | 42 (6) | 1.4 (0.9–2.3) | 17 (4.2) | 0.9 (0.5–1.6) | 16 (8.8) | 2.0 (1.1–3.8) | |
| Cirrhosis | No | 679 (98.7) | 644 (98.9) | 1.0 | 407 (99.0) | 1.0 | 178 (98.3) | 1.0 |
| Yes | 9 (1.3) | 7 (1.1) | 0.7 (0.3–2.0) | 4 (1.0) | 0.7 (0.2–2.2) | 3 (1.7) | 1.2 (0.3–4.6) | |
| Mononucleosis | No | 634 (92.7) | 600 (92.0) | 1.0 | 380 (92.7) | 1.0 | 168 (91.8) | 1.0 |
| Yes | 50 (7.3) | 52 (8.0) | 1.2 (0.8–1.8) | 30 (7.3) | 1.0 (0.6–1.7) | 15 (8.2) | 1.2 (0.6–2.0) | |
| Chronic fatigue syndrome | No | 682 (99.4) | 642 (99.1) | 1.0 | 409 (99.8) | 1.0 | 175 (97.8) | 1.0 |
| Yes | 4 (0.6) | 6 (0.9) | 1.6 (0.4–5.8) | 1 (0.2) | 0.4 (0.0–4.0) | 4 (2.2) | 4.2 (1.0–18)d | |
| Epilepsy | No | 679 (99.0) | 644 (98.6) | 1.0 | 406 (98.8) | 1.0 | 181 (98.9) | 1.0 |
| Yes | 7 (1.0) | 9 (1.4) | 1.4 (0.5–3.7) | 5 (1.2) | 1.2 (0.4–3.8) | 2 (1.1) | 1.1 (0.2–5.5) | |
Medical conditions diagnosed within 1 year of leukemia diagnosis (cases) or 1 year of recruitment date (controls) were excluded Bold values represent statistical significance at the p < 0.05 level
Models adjusted for age (deciles), sex, and BMI (underweight/normal weight, overweight, obese)
The number of cases and controls for each condition does not add up to the total number of cases (n = 670) and controls (n = 701) due to missing data for the particular medical condition
Includes the following conditions: polymyositis, rheumatoid arthritis, Sjogren’s syndrome, systemic lupus erythematosus, hypothyroidism, hyperthyroidism, ankylosing spondylitis, Crohn’s disease, ulcerative colitis, and celiac disease
Heart disease and chronic fatigue syndrome are rounded to 1.0 for the lower CI; thus, they are not significant at α < 0.05
Table 3.
Associations between personal history of cancer/cancer treatment(s) and risk of myeloid leukemia
| Controls | Myeloid leukemia |
AML |
CML |
|||||
|---|---|---|---|---|---|---|---|---|
| N (%) | N (%) | OR (95% CI) | N (%) | OR (95% CI) | N (%) | OR (95% CI) | ||
| Personal history of cancera,b | ||||||||
| No | 645 (95.8) | 557 (92.1) | 1.0 | 353 (93.1) | 1.0 | 152 (89.4) | 1.0 | |
| Yes | 50 (7.2) | 106 (15.8) | 2.1 (1.3–3.5) | 65 (16) | 1.8 (1.0–3.1) | 32 (17.2) | 3.6 (1.9–7.0) | |
| Radiation and/or chemotherapy treatmentb,c | ||||||||
| None | 673 (96.4) | 605 (90.6) | 1.0 | 379 (90.5) | 1.0 | 170 (91.9) | 1.0 | |
| Radiation alone | 12 (1.7) | 18 (2.7) | 1.6 (0.8–3.4) | 9 (2.2) | 1.3 (0.6–3.2) | 7 (3.8) | 2.7 (1.0–7.3) | |
| Chemotherapy alone |
4 (0.6) | 15 (2.3) | 4.1 (1.3–12.6) | 11 (2.6) | 5.1 (1.6–16.5) | 4 (2.2) | 3.6 (0.9–15.1) | |
| Both | 6 (0.9) | 25 (3.7) | 5.0 (2.0–12.3) | 19 (4.5) | 6.0 (2.4–15.4) | 3 (1.6) | 2.2 (0.5–9.5) | |
Bold values represent statistical significance at the p < 0.05 level
ORs adjusted for age (deciles), sex, and BMI (underweight/normal weight, overweight, obese), radiation or chemotherapy treatment (yes vs. no)
Three controls and five cases each had missing data on personal history of cancer, radiation treatment, and chemotherapy treatment
ORs adjusted for age (deciles), sex, and BMI (underweight/normal weight, overweight, obese)
Results
Case and control frequencies were similar across age categories, the matching variable (Table 1). Cases were more likely to be male and less likely to be non-Hispanic White. CML cases were more likely to report having less than a high school education and an income <$40,000 per year. Both AML and CML cases were more likely to report ever having smoked, being overweight or obese, and having been exposed to benzene or other solvents.
Table 1.
Characteristics of study participants
| Characteristic | Controls (N = 701) No. (%) |
Myeloid leukemia (N = 670) |
AML (N = 420) No. (%) |
CML (N = 186) No. (%) |
|---|---|---|---|---|
| Age (years) a | ||||
| 20–29 | 45 (6.4) | 42 (6.3) | 26 (6.2) | 12 (6.5) |
| 30–39 | 53 (7.6) | 48 (7.2) | 32 (7.6) | 15 (8.1) |
| 40–49 | 102 (14.6) | 102 (15.2) | 64 (15.2) | 33 (17.7) |
| 50–59 | 157 (22.4) | 143 (21.3) | 90 (21.4) | 48 (25.8) |
| 60–69 | 205 (29.2) | 195 (29.1) | 129 (30.7) | 48 (25.8) |
| 70–79 | 139 (19.8) | 140 (20.9) | 79 (18.8) | 30 (16.1) |
| Sex | ||||
| Male | 343 (48.9) | 392 (58.5) | 249 (59.3) | 107 (57.5) |
| Female | 358 (51.1) | 278 (41.5) | 171 (40.7) | 79 (42.5) |
| Race/ethnicity | ||||
| White, non- Hispanic |
671 (95.7) | 629 (93.9) | 394 (93.8) | 172 (92.5) |
| Other | 30 (4.3) | 41 (6.1) | 26 (6.2) | 14 (7.5) |
| Education | ||||
| ≤High school graduate |
231 (33.0) | 212 (31.6) | 122 (29.1) | 67 (36.0) |
| Some post-high school |
240 (34.2) | 236 (35.2) | 161 (38.3) | 55 (29.6) |
| College graduate |
230 (32.8) | 222 (33.1) | 137 (32.6) | 64 (34.4) |
| Income (US dollars) b | ||||
| <$40,000 | 255 (36.9) | 260 (39.7) | 149 (36.4) | 80 (43.7) |
| $40,000– $80,000 |
281 (40.7) | 246 (37.6) | 165 (40.3) | 64 (35.0) |
| >$80,000 | 155 (22.4) | 149 (22.8) | 95 (23.2) | 39 (21.3) |
| Smoking status | ||||
| Never | 358 (51.7) | 288 (43.4) | 179 (43.2) | 81 (43.8) |
| Ever | 334 (48.3) | 375 (56.6) | 235 (56.8) | 104 (56.2) |
| Married | ||||
| Yes | 513 (73.2) | 481 (71.8) | 315 (75.0) | 122 (65.6) |
| No | 188 (26.8) | 189 (28.2) | 105 (25.0) | 64 (34.4) |
| BMI category c | ||||
| Normal | 227 (32.5) | 152 (22.8) | 96 (22.9) | 41 (22.2) |
| Overweight | 242 (34.7) | 231 (34.6) | 147 (35.1) | 59 (31.9) |
| Obese | 229 (32.8) | 285 (42.7) | 176 (42.0) | 85 (46.0) |
| History of Benzene or other solvent exposure | ||||
| No | 651 (93.3) | 560 (84.0) | 343 (82.1) | 164 (88.2) |
| Yes | 47 (6.7) | 107 (16.0) | 75 (17.9) | 22 (11.8) |
Matching variable
Income may include retirement income
BMI based on weight 2 years prior to diagnosis/recruitment date
No significant associations were found between myeloid leukemia overall, AML, or CML, and any autoimmune conditions except for UC (Table 2). There was an increased risk of myeloid leukemia associated with UC (OR = 3.5; 95% CI, 1.1–11), which was stronger for AML (OR = 3.8; 95% CI, 1.1–13) than for CML (OR = 2.0; 95% CI, 0.4–11). For CML, we observed a significant increased risk of peptic ulcer (OR = 2.0; 95% CI, 1.1–3.8). Adjusting for regular aspirin use did not materially change the risk estimate (data not shown).
We also conducted more stringent analyses that excluded individuals who reported that their medical condition occurred within 3 years of their leukemia diagnosis or questionnaire completion date for controls. Stronger associations, although less precise, were observed between UC and myeloid leukemia overall (OR = 4.5; 95% CI, 1.2–16.4) and for AML (OR = 5.2; 95% CI, 1.3–20.1) but not for CML (OR = 1.3; 95% CI, 0.1–13.5). For peptic ulcer, similar results were obtained to the results reported in Table 2 for myeloid leukemia overall (OR = 1.5; 95% CI, 0.9–2.4), AML (OR = 0.9; 95% CI, 0.5–1.7), and CML (OR = 2.0; 95% CI, 1.0–3.8).
A personal history of cancer increased the risk of myeloid leukemia overall (OR = 2.6; 95% CI, 1.8–3.7), AML (OR = 2.6; 95% CI, 1.7–3.9), and CML (OR = 3.5; 95% CI, 2.0–5.8) (Table 3). The associations remained significant after adjusting for previous radiation or chemotherapy treatment for myeloid leukemia overall (OR = 2.1; 95% CI, 1.3–3.5), AML (OR = 1.8; 95% CI, 1.0–3.1), and CML (OR = 3.6; 95% CI, 1.9–7.0). Exclusion of reported prior hematopoietic malignancies did not materially change results (data not shown). In models that excluded individuals who reported a previous cancer diagnosis within 5 years or their leukemia diagnosis or questionnaire completion date for controls, associations remained significant for myeloid leukemia overall (OR = 2.8; 95% CI, 1.8–4.4), AML (OR = 2.9; 95% CI, 1.8–4.7), and CML (OR = 2.8; 95% CI, 1.4–5.5). Radiation treatment only for a prior cancer significantly increased the risk of CML (OR = 2.7; 95% CI, 1.0–7.3) only. Previous chemotherapy treatment was associated with stronger risks than radiation treatment for myeloid leukemia overall (OR = 4.1; 95% CI, 1.3–12.6), AML (OR = 5.1; 95% CI, 1.6–16.5), and CML (OR = 3.6; 95% CI, 0.9–15.1). Past cancer treatment with both radiation and chemotherapy increased the risk of myeloid leukemia overall by fivefold (95% CI, 2.0–12.3), of AML by sixfold (95% CI, 2.4–15.4), and of CML by 2.2-fold (95% CI, 0.5–9.5) (Table 3). No particular pattern emerged for types of cancers reported by cases versus controls, with prostate and breast cancer reported most frequently by men and women, respectively, in both groups (data not shown).
Discussion
A few population-based studies have examined the associations between UC and adult myeloid leukemia. In a Swedish cohort of 27,559 patients with UC, approximately twofold increased risks for both AML and CML were reported [4]. Another Swedish study of UC patients reported a significant standardized incident ratio (SIR) for CML (SIR, 3.0; 95% CI, 1.4–5.3), but not AML, for cases hospitalized at least 1 year before their leukemia diagnosis [5]. A US study of adult leukemia (n = 7924 AML and 2174 CML cases) using data from the Surveillance Epidemiology and End Results Medicare database reported a significant increased risk of AML (OR = 1.7), but not CML (OR = 0.7), associated with UC [1]. Finally, a large 2011 Swedish study that linked the cancer and inpatient registries reported no association between AML and UC (OR = 0.8) [6]. However, data on UC were obtained from the inpatient patient registry; thus, only severe cases requiring hospitalization would be captured. We used a case–control study design that collected data through survey methods, which strengthen the existing evidence for a positive relationship between UC and myeloid leukemia, particularly AML.
A relationship between inflammatory bowel diseases (UC and Crohn’s disease) and leukemia may result from thiopurine treatment [10]. Azathioprine, an immune suppressant, has been classified as a carcinogen for non-Hodgkin lymphoma and skin cancer [11]. An increased frequency of somatic mutations in the hypoxanthine phosphoribosyltransferase (HPRT) locus has been detected in the T lymphocytes of thiopurine-treated inflammatory bowel disease patients [12]. Interestingly, a Nordic study of childhood leukemia survivors reported significant associations between duration and intensity of 6-MP/Methotrexate maintenance therapy and development of secondary AML [13]. However, leukemia cases have also been noted in association with UC in the absence of thiopurine treatment [14], perhaps due to coincidence or to underlying biological factors predisposing to both conditions. It will be important in future studies to collect information on medical treatment for UC and Crohn’s disease to determine whether thiopurines or other drugs used to treat UC and Crohn’s disease increase the risk of myeloid leukemia.
Patients with myeloproliferative disorders have been reported as having a higher prevalence of gastric ulcers [15], but there are no known reported associations between peptic ulcer and CML, which is classified as a myeloproliferative neoplasm. Gastric ulcers are associated with aspirin use and Heliobacter (H.) pylori infection, both inducing stomach lining inflammation. Aspirin use was not a confounder in our study. H. pylori infection has not been linked to CML specifically, although it has been linked with other adult hematologic malignancies [16] and with childhood leukemia [17], with the underlying biology remaining unclear.
Three studies [1, 6, 7] found significant positive associations between rheumatoid arthritis and myeloid leukemia, which we did not confirm.
Finally, a previous cancer diagnosis and treatment with radiation or chemotherapy were strongly related to both AML and CML. Cancer treatment with radiation, alkylating agents, and DNA topoisomerase II inhibitors has been linked to therapy-related AML (t-AML) [18]. Both AML and CML have been associated with radiation exposure in studies of atomic bomb survivors [19, 20], and evidence also suggests an association with previous cancer radiation treatment [21–23]. We note that in our study, previous radiation treatment for cancer was not significantly associated with AML, which is in contrast to studies of atomic bomb survivors that have indicated an age of exposure-dependent increased risk of AML in association with radiation exposure. The differences between our results and those of atomic bomb survivors could be due to age at exposure since no significant effect of radiation exposure on AML risk was observed in individuals who were older than 30 years at the time of the atomic bombings [20]. Evidence also indicates an increased risk of CML in association with cytotoxic therapy [18, 24]. It is noteworthy that a personal history of cancer was an independent predictor for both AML and CML in models that controlled for therapy. One possible explanation for this finding is that genetic factors that influence susceptibility to multiple cancers may be involved.
Strengths of our study include a large population-based study design with data collected from patients directly, rather than from surrogate respondents. Participation rates and sociodemographic factors were also similar for cases and controls; nevertheless, selection bias could have influenced our results if controls who participated in the study were less likely to have medical conditions than individuals in the source cohort. It is also possible that our findings could be due to chance, given multiple comparisons and small numbers of exposed subjects. Cases may be more likely than controls to remember past medical conditions. Differential reporting may also have occurred due to symptoms (such as UC) varying over time [25]. Notably, UC prevalence among controls was higher than that expected from previous estimates [26]. Finally, generalizability is limited to adult myeloid leukemia patients with relatively longer survival times.
In summary, we found significant increased risks between a history of certain medical conditions and adult leukemia. UC significantly increased the risk of AML, while for CML, peptic ulcer significantly increased the risk. A personal history of cancer strongly increased the risk of AML and especially CML with our estimates indicating that chemotherapy was a stronger risk factor than radiation therapy. The increased risk of myeloid leukemia in association with a personal history of cancer that is independent from therapy suggests that some individuals with leukemia may have an underlying genetic susceptibility to cancer.
Acknowledgments
The authors acknowledged the grants from NIH T32 CA099936, R25 CA047888, R01 CA107143, and K05 CA157439.
Footnotes
Conflict of interest The authors have no financial disclosures to report.
Contributor Information
Kimberly J. Johnson, Brown School, Washington University in St. Louis, St. Louis, MO 63130, USA; Department of Pediatrics, School of Medicine, Washington University in St. Louis, St. Louis, MO 63110, USA
Cindy M. Blair, Department of Nutrition Sciences, University of Alabama at Birmingham, Birmingham, AL 35294, USA
James M. Fink, Department of Laboratory Medicine and Pathology, Hennepin Faculty Associates, Minneapolis, MN 55415, USA
James R. Cerhan, Division of Epidemiology, Department of Health Sciences Research, Mayo Clinic, Rochester, MN 55905, USA
Michelle A. Roesler, Division of Epidemiology and Clinical Research, Department of Pediatrics, University of Minnesota, MMC 422, 420 Delaware St. S.E., Minneapolis, MN 55455, USA
Betsy A. Hirsch, Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, MN 55455, USA
Phuong L. Nguyen, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN 55905, USA
Julie A. Ross, Division of Epidemiology and Clinical Research, Department of Pediatrics, University of Minnesota, MMC 422, 420 Delaware St. S.E., Minneapolis, MN 55455, USA; Masonic Cancer Center, University of Minnesota, Minneapolis, MN 55455, USA
References
- 1.Anderson LA, Pfeiffer RM, Landgren O, Gadalla S, Berndt SI, Engels EA. Risks of myeloid malignancies in patients with autoimmune conditions. Br J Cancer. 2009;100(5):822–828. doi: 10.1038/sj.bjc.6604935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Karlen P, Lofberg R, Brostrom O, Leijonmarck CE, Hellers G, Persson PG. Increased risk of cancer in ulcerative colitis: a population-based cohort study. Am J Gastroenterol. 1999;94(4):1047–1052. doi: 10.1111/j.1572-0241.1999.01012.x. [DOI] [PubMed] [Google Scholar]
- 3.Winther KV, Jess T, Langholz E, Munkholm P, Binder V. Long-term risk of cancer in ulcerative colitis: a population-based cohort study from Copenhagen County. Clin Gastroenterol Hepatol. 2004;2(12):1088–1095. doi: 10.1016/s1542-3565(04)00543-9. [DOI] [PubMed] [Google Scholar]
- 4.Askling J, Brandt L, Lapidus A, Karlen P, Bjorkholm M, Lofberg R, et al. Risk of haematopoietic cancer in patients with inflammatory bowel disease. Gut. 2005;54(5):617–622. doi: 10.1136/gut.2004.051771. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Hemminki K, Li X, Sundquist J, Sundquist K. Cancer risks in ulcerative colitis patients. Int J Cancer. 2008;123(6):1417–1421. doi: 10.1002/ijc.23666. [DOI] [PubMed] [Google Scholar]
- 6.Kristinsson SY, Bjorkholm M, Hultcrantz M, Derolf AR, Landgren O, Goldin LR. Chronic immune stimulation might act as a trigger for the development of acute myeloid leukemia or myelodysplastic syndromes. J Clin Oncol. 2011;29(21):2897–2903. doi: 10.1200/JCO.2011.34.8540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Askling J, Fored CM, Baecklund E, Brandt L, Backlin C, Ekbom A, et al. Haematopoietic malignancies in rheumatoid arthritis: lymphoma risk and characteristics after exposure to tumour necrosis factor antagonists. Ann Rheum Dis. 2005;64(10):1414–1420. doi: 10.1136/ard.2004.033241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ross JA, Blair CK, Cerhan JR, Soler JT, Hirsch BA, Roesler MA, et al. Nonsteroidal anti-inflammatory drug and acetaminophen use and risk of adult myeloid leukemia. Cancer Epidemiol Biomarkers Prev. 2011;20(8):1741–1750. doi: 10.1158/1055-9965.EPI-11-0411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Slattery ML, Edwards SL, Caan BJ, Kerber RA, Potter JD. Response rates among control subjects in case–control studies. Ann Epidemiol. 1995;5(3):245–249. doi: 10.1016/1047-2797(94)00113-8. [DOI] [PubMed] [Google Scholar]
- 10.Das KK, Nishino HT, Chan AT. Treatment-associated acute myeloid leukemia in a patient with Crohn’s disease on 6-mercaptopurine. Inflamm Bowel Dis. 2010;16(9):1454–1456. doi: 10.1002/ibd.21205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Grosse Y, Baan R, Straif K, Secretan B, El Ghissassi F, Bouvard V, et al. A review of human carcinogens-Part A: pharmaceuticals. Lancet Oncol. 2009;10(1):13–14. doi: 10.1016/s1470-2045(08)70286-9. [DOI] [PubMed] [Google Scholar]
- 12.Nguyen T, Vacek PM, O’Neill P, Colletti RB, Finette BA. Mutagenicity and potential carcinogenicity of thiopurine treatment in patients with inflammatory bowel disease. Cancer Res. 2009;69(17):7004–7012. doi: 10.1158/0008-5472.CAN-09-0451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Schmiegelow K. Epidemiology of therapy-related myeloid neoplasms after treatment for pediatric acute lymphoblastic leukemia in the nordic countries. Mediterr J Hematol Infect Dis. 2011;3(1):e2011020. doi: 10.4084/MJHID.2011.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Fabry TL, Sachar DB, Janowitz HD. Acute myelogenous leukemia in patients with ulcerative colitis. J Clin Gastroenterol. 1980;2(3):225–227. doi: 10.1097/00004836-198009000-00003. [DOI] [PubMed] [Google Scholar]
- 15.Karaoglu AO, Kadikoylu G, Yukselen V, Yasa MH, Bolaman Z. Gastrointestinal lesions and Helicobacter pylori in patients with myeloproliferative disorders. Saudi Med J. 2004;25(12):1913–1916. [PubMed] [Google Scholar]
- 16.Grulich AE, Vajdic CM. The epidemiology of non-Hodgkin lymphoma. Pathology. 2005;37(6):409–419. doi: 10.1080/00313020500370192. [DOI] [PubMed] [Google Scholar]
- 17.Lehtinen M, Ogmundsdottir HM, Bloigu A, Hakulinen T, Hemminki E, Gudnadottir M, et al. Associations between three types of maternal bacterial infection and risk of leukemia in the offspring. Am J Epidemiol. 2005;162(7):662–667. doi: 10.1093/aje/kwi261. [DOI] [PubMed] [Google Scholar]
- 18.Schottenfeld D, Fraumeni JF. Cancer epidemiology and prevention. 3rd edn Oxford University Press; Oxford: 2006. [Google Scholar]
- 19.Preston DL, Kusumi S, Tomonaga M, Izumi S, Ron E, Kuramoto A, et al. Cancer incidence in atomic bomb survivors. Part III. Leukemia, lymphoma and multiple myeloma, 1950–1987. Radiat Res. 1994;137(2 Suppl):S68–S97. [PubMed] [Google Scholar]
- 20.Tsushima H, Iwanaga M, Miyazaki Y. Late effect of Atomic bomb radiation on myeloid disorders: leukemia and myelodysplastic syndromes. Int J Hematol. 2012;95(3):232–238. doi: 10.1007/s12185-012-1002-4. [DOI] [PubMed] [Google Scholar]
- 21.Howard RA, Gilbert ES, Chen BE, Hall P, Storm H, Pukkala E, et al. Leukemia following breast cancer: an international population-based study of 376,825 women. Breast Cancer Res Treat. 2007;105(3):359–368. doi: 10.1007/s10549-006-9460-0. [DOI] [PubMed] [Google Scholar]
- 22.Wang KL, Lin LY, Chen PM, Lin HD. Chronic myeloid leukemia after treatment with 131 for thyroid carcinoma. J Chin Med Assoc. 2005;68(5):230–233. doi: 10.1016/s1726-4901(09)70213-8. [DOI] [PubMed] [Google Scholar]
- 23.Bauduer F, Ducout L, Dastugue N, Marolleau JP. Chronic myeloid leukemia as a secondary neoplasm after anti-cancer radiotherapy: a report of three cases and a brief review of the literature. Leuk Lymphoma. 2002;43(5):1057–1060. doi: 10.1080/10428190290021533. [DOI] [PubMed] [Google Scholar]
- 24.Tsuzuki M, Handa K, Yamamoto K, Hasegawa A, Yamamoto Y, Watanabe M, et al. Chronic myeloid leukemia following chemotherapy with 5′-deoxy-5-fluorouridine for gastric cancer. Intern Med. 2008;47(19):1739–1741. doi: 10.2169/internalmedicine.47.1072. [DOI] [PubMed] [Google Scholar]
- 25.Health Pm. Ulcerative Colitis. [cited 2011 11/4/2011]. 2011.
- 26.Loftus EV., Jr Clinical epidemiology of inflammatory bowel disease: Incidence, prevalence, and environmental influences. Gastroenterology. 2004;126(6):1504–1517. doi: 10.1053/j.gastro.2004.01.063. [DOI] [PubMed] [Google Scholar]