Among patients with systemic lupus erythematosus (SLE), cancer risks overall were associated with older age at diagnosis and male sex, and lung cancer specifically was related to smoking, analysis of data from a large multicenter cohort found.
In a multivariable analysis, the hazard ratio for any type of cancer among patients enrolled in the Systemic Lupus International Collaborating Clinics Inception Cohort was 1.05 (95% CI 1.03-1.06) for greater age at baseline, while the risk for women was lower, with a hazard ratio of 0.47 (95% CI 0.26-0.85), reported Sasha Bernatsky, MD, of McGill University in Montreal, and colleagues.
As shown in their study online in , for lung cancer specifically a history of smoking more than 15 cigarettes per day was associated with an almost seven-fold increased risk (HR 6.64, 95% CI 1.43-30.9).
Interest has been growing with regard to the risk of cancer among patients with autoimmune diseases, including lupus, because of a potential role of inflammation in malignancy development. In addition, certain medications, such as cyclophosphamide, have been linked with cancer in the past.
Previous studies of cancer risk in SLE patients have been limited in size and a reliance on retrospective administrative data, which do not reflect various clinical factors and drug exposures.
To address this knowledge gap, the researchers analyzed data for 1,668 patients enrolled in the cohort from 1999 to 2011, who were seen yearly with information collected on disease activity, medication usage, and cancer diagnoses.
During follow-up, which totaled 15,014 person-years (mean 9 years per patient), there were 65 cancers -- 15 breast cancers, 10 non-melanoma skin cancers, seven lung cancers, six prostate cancers, six hematologic malignancies, five melanomas, three cervical cancers, three kidney cancers, two gastric cancers, two head and neck cancers, two thyroid cancers, and one each of rectal cancer, sarcoma, thymoma, and uterine cancer.
When the researchers compared the incident cancers according to baseline patient characteristics, the results showed that whereas 89.3% of patients without cancer were women, only 78.3% of those who developed cancer were.
Mean age at the time of SLE diagnosis was 34.2 in the no-cancer group compared with 45.6 in the cancer group. Current or past smoking was reported in 33.3% of the no-cancer group but in 47.7% of the cancer group. White race was reported for 48.7% of the no-cancer group vs 67.7% of the cancer group.
In an unadjusted univariate analysis, factors positively associated with any type of cancer were older age at SLE diagnosis (HR 1.06, 95% CI 1.04-1.07), white race (HR 2.24, 95% CI 1.33-3.78), and ever smoking (HR 1.72, 95% CI 1.06-2.80), while female sex was associated with a lower risk (HR 0.35, 95% CI 0.20-0.60). However, in the multivariate analysis, only older age and female sex remained significant.
The researchers then estimated risks for the specific types of cancer. For breast cancer, only older age at the time of diagnosis showed a significantly increased risk in the multivariate analysis (HR 1.06, 95% CI 1.02-1.10), while antimalarial use was associated with a lower risk (HR 0.28, 95% CI 0.09-0.90).
For non-melanoma skin cancer, associations were observed for older age at diagnosis (HR 1.06, 95% CI 1.02-1.11) and cyclophosphamide use (HR 15.3, 95% CI 3.03-77.5). "It is well known that non-melanoma skin cancers may be triggered by immunosuppressive drugs," the researchers explained, adding that the effects of drugs on non-melanoma skin cancers in SLE also can be complicated by the photosensitivity typical of the disease.
For lung cancer, all cases occurred in patients who were current or past smokers. The risk was lower among women (HR 0.18, 95% CI 0.04-0.86) and higher among those who were heavier smokers.
None of the lung cancer patients had been treated with cyclophosphamide or methotrexate, but all had received long-term treatment with antimalarials; it was not possible, therefore, to provide risk estimates for those medications, the team noted.
With hematologic cancers, once again older age at SLE diagnosis had a significant association (HR 1.06, 95% CI 1-1.13), as did having ever been in the highest quartile of SLE disease activity (HR 7.14, 95% CI 1.13-45.3). All patients with these malignancies were white and had ever smoked but none had been treated with cyclophosphamide, so risk estimates for those factors also could not be calculated.
One major finding was that certain cancers, particularly lung cancer, occurred more often in patients who had ever smoked. This was a key modifiable risk factor, the researchers pointed out.
The observation that male sex and older age at the time of SLE diagnosis were associated with higher risks in most types of cancer "may be, at least in part, because these demographic groups are at greater cancer risk in the general population," Bernatsky and colleagues wrote.
"However, further study of cancer risk in the potentially vulnerable SLE populations would be of interest, to determine if longer windows of observation result in the same findings and/or identify any additional risk factors," the researchers added.
A limitation of the study, they said, was its observational design, which cannot establish causality.
Disclosures
The researchers received support from the National Institute for Health Research, Arthritis Research UK, the National Institutes of Health, the National Health Service Trust, the Danish Rheumatism Association, the CDC, the Basque Government, and the Singer Family Fund for Lupus Research.
Bernatsky reported no potential competing interests; co-authors reported financial relationships with Bristol Myers Squibb, GlaxoSmithKline, Eli Lilly, Pfizer, UCB, AbbVie, AstraZeneca, Biotest, Celgene, Janssen, Novartis, Pfizer, Servier, and Exagen Diagnostics.
Primary Source
Arthritis Care & Research
Bernatsky S, et al "Cancer risk in a large inception SLE cohort: Effects of demographics, smoking, and medications" Arthritis Care Res 2020; DOI: 10.1002/acr.24425.