Key points

Introduction

Systemic lupus erythematosus (SLE), an autoimmune disorder with complex environmental and genetic interactions, affects approximately 1 in 1000 women in North America. In the last 5 decades, advances in management have improved 5-year survival to more than 90%. The longer survival translates, in some cases, to considerable long-term morbidity, including distinct cancer risk profiles.

For more than a decade, autoimmune disorders such as SLE have been associated with an increased cancer risk, particularly for certain cancer types. However, the underlying pathophysiologic mechanisms are still not fully understood. In SLE, the proposed pathways currently include a possible link with medications, inherent immune system abnormalities, an overlap with clinical syndromes such as Sjögren, viral infections, as well as traditional lifestyle cancer risk factors. The most recent evidence shows a slight (standard incidence ratio [SIR], 1.14; 95% confidence interval [CI], 1.05–1.23) increase in cancer risk overall. However, considerable variation between specific cancer subtypes occurs within these data. The risk of hematologic cancers, especially non-Hodgkin lymphoma (NHL), is increased about 3-fold, although breast, endometrial, and possibly prostate and ovarian cancers seem to be associated with a decreased risk in the SLE population.

Introduction

Systemic lupus erythematosus (SLE), an autoimmune disorder with complex environmental and genetic interactions, affects approximately 1 in 1000 women in North America. In the last 5 decades, advances in management have improved 5-year survival to more than 90%. The longer survival translates, in some cases, to considerable long-term morbidity, including distinct cancer risk profiles.

For more than a decade, autoimmune disorders such as SLE have been associated with an increased cancer risk, particularly for certain cancer types. However, the underlying pathophysiologic mechanisms are still not fully understood. In SLE, the proposed pathways currently include a possible link with medications, inherent immune system abnormalities, an overlap with clinical syndromes such as Sjögren, viral infections, as well as traditional lifestyle cancer risk factors. The most recent evidence shows a slight (standard incidence ratio [SIR], 1.14; 95% confidence interval [CI], 1.05–1.23) increase in cancer risk overall. However, considerable variation between specific cancer subtypes occurs within these data. The risk of hematologic cancers, especially non-Hodgkin lymphoma (NHL), is increased about 3-fold, although breast, endometrial, and possibly prostate and ovarian cancers seem to be associated with a decreased risk in the SLE population.

Hematologic cancers

An increased incidence of hematologic malignancies in the SLE population was initially suggested more than 3 decades ago and has now been supported by many studies. Based on incidence and mortality data generated from the large, multicentre, international SLE cohort contributed by Systemic Lupus International Collaborating Clinics (SLICC) and other investigators, it was observed that NHL incidence (SIR, 3.02; 95% CI, 2.48–3.63) and mortality (standardized mortality ratio [SMR], 2.8; 95% CI, 1.2–5.6) risks were particularly increased in patients with SLE, compared with expected general population cancer rates. Further studies identified an increased risk for Hodgkin lymphomas (HL) and leukemia in patients with SLE. One study suggests that the risk of developing multiple myeloma might also be increased in patients with SLE, but additional studies are required to confirm this association.

Genetic changes, such as the t(14:18) translocation (which results in overexpression of BCL2 caused by juxtaposition of the BCL2 gene next to an immunoglobulin gene) have been linked to lymphoma development. In addition, a polymorphism involving the interleukin (IL)-1 receptor antagonist has been noted to be significantly overexpressed in a group of patients with secondary acute myeloid leukemia.

SLE activity has been invoked as a potential factor to explain the increased lymphoma risk in SLE. Although increased disease activity has been associated with higher cancer risk in certain autoimmune disorders (for example, in rheumatoid arthritis [RA], higher disease activity was shown to be a marker of lymphoma risk, in one study), the immune system is also responsible for targeting abnormal cells. Thus, the effects of disease activity in one rheumatic disease, such as RA, may be different in another, such as SLE. Case-cohort analyses of lymphoma cases in the large international multicentre SLE cohort revealed no clear relationship between disease activity and lymphoma risk (hazard ratio [HR], 0.68; 95% CI, 0.36–1.29).

Diffuse large B-cell lymphoma (DLBCL) is the most increased subtype of lymphoma in SLE. DLBCL lesions arise from activated lymphocytes, the cell line responsible for most of the inflammation in autoimmune disorders such as SLE, which suggests that the chronic inflammatory state seen in patients with SLE contributes to the cancer risk profile in SLE. Furthermore, DLBCL lesions that develop in patients with SLE seem to highly express APRIL (a proliferation-inducing ligand), a cytokine from the tumor necrosis factor (TNF) ligand superfamily that is essential for B-cell survival and development. The investigators of that study proposed that the overproduction of APRIL might mediate the development of lymphoma in SLE and other rheumatic diseases. The mechanism by which APRIL induces lymphoma is unclear, but it seems to play a role in allowing NHL B cells to escape apoptosis.

The Epstein-Barr virus (EBV) is suggested to have a role in the pathophysiology of SLE, and may also contribute to the increased risk of malignancy in SLE. EBV seropositivity is only slightly increased in patients with SLE compared with the general population, but some data indicate altered ability to clear this viral infection in SLE. Increased viral load, EBV mRNA expression, EBV-directed antibodies, and decreased EBV-directed cell immunity have all been shown in patients with SLE compared with healthy controls. This finding is relevant to studies of the association of SLE and cancer because EBV has been associated with malignancies in certain populations. For example, EBV is associated with neoplastic processes such as Burkitt lymphoma, Hodgkin disease, immunosuppression-related DLBCL, and nasopharyngeal and gastric carcinomas. In addition, EBV viral load was shown to be prognostic of impending lymphoproliferative disorder in transplant recipients. The mechanism by which EBV is thought to promote malignancy seems to involve B-cell immortalization, manipulating host chromatin-remodeling machinery, and promoting cell migration and resistance to apoptosis through p53 blc-2 A20 and Fas modulation.

Polymorphism of the Fas gene and an epistatic relationship between Fas and Sle1 genes has been reported in the SLE population and altered Fas-mediated apoptosis contributes to the pathophysiology of autoimmune lymphoproliferative syndrome (ALPS). ALPS is a rare, autosomal dominant, inherited genetic disorder that involves defective apoptosis of lymphocytes and subsequent increased risk of developing both autoimmunity (including many SLE manifestations: rashes, nephritis, arthritis, and autoantibodies) and lymphomas. Thus, the family history of affected persons features multiple manifestations of both autoimmune disease and lymphoma. The exact prevalence of ALPS is uncertain, but about 100 affected individuals have been documented worldwide. Although this condition is an example of how a genetic defect can increase an individual’s risk for both SLE and lymphoma, this specific autosomal dominant inherited genetic disorder is too rare to explain most of the lymphoma cases that arise in SLE.

In addition, a link between primary Sjögren syndrome and lymphoma has been shown in many studies, and it has thus been suggested that lupus-induced secondary Sjögren syndrome could account for the heightened risk of hematologic malignancies. Our case-cohort analyses based on the large multicentre international SLE cohort showed a trend supporting an increased risk for lymphoma development in patients with Sjögren syndrome (HR, 1.79; 95% CI, 0.88–3.62). Here, Sjögren syndrome was based on clinical judgment, as opposed to requiring patients to fulfill specific criteria, and thus may have been subject to nondifferential misclassification of the exposure, which could have biased this result toward the null value. Thus, it remains possible that secondary Sjögren syndrome may explain some (but not all) of the increased lymphoma risk in SLE.

There continues to be great interest in the potential role of medications in mediating lymphoma risk in SLE. Case-cohort analysis of the multicentre international SLE cohort suggested an increased risk of lymphoma in subjects exposed to cyclophosphamide (SIR, 2.80; 95% CI, 0.87–8.98) as well as to cumulative glucocorticoids (SIR, 2.57; 95% CI, 0.94–7.04) compared with SLE controls without cancer. The mechanism of cyclophosphamide’s effects in increasing cancer risk might be direct, such as through its immunosuppression and cytotoxic properties, or indirect, through, for example, promoting the emergence of oncogenic virus (eg, EBV or human papilloma virus [HPV]).

Lung cancers

Lung cancer also has a reported increased incidence (SIR, 1.30; 95% CI, 1.04–1.60) and mortality (SMR, 2.3; 95% CI, 1.6–3.0) in the SLE population. Although a potential trend toward overrepresentation of rarer types (including bronchoalveolar and carcinoid) was identified in one study, the overall histologic distribution was comparable with that of the general population.

Genetic associations are an interesting possible explanation for the association between SLE and lung cancer, because there is evidence of shared susceptibility loci (4p15.1–15.3 and 6p21). Although the idea of a genetic link predisposing to both lung cancer and SLE is plausible, it remains premature and thus this represents an area for future research.

The proven link between lung cancer, fibrosis, and inflammation has also led to the consideration of pulmonary involvement (that is, alveolitis and/or fibrosis) in SLE as playing a direct role in lung cancer risk. However, in our large, multicentre, international SLE cohort, clear documentation of preexisting pulmonary damage was reported in only 1 lung case; however, more prospective analyses are warranted.

Smoking may represent a shared environmental risk factor between both lung cancer and SLE. Data support an association between lupus and smoking, which might contribute to the reported increased lung cancer incidence in SLE. Case-cohort analyses have strongly suggested smoking as an important predictor of lung cancer in SLE. In addition, only 20% of patients affected by lung cancer from the large international SLE cohort had previously been exposed to immunotherapy ; this may suggest that drugs are not the primary cause of lung cancer in patients with SLE.

Cervical cancers and dysplasia

It has been suggested that women with SLE are at an increased risk for invasive cervical cancer and for cervical dysplasia ; however, not all analyses to date have been definitive. The multicentre international SLE cohort, which excluded cervical dysplasia and carcinoma in situ, was unable to document an increased risk of invasive cervical cancer in SLE, although a trend was evident (SIR, 1.27; 95% CI, 0.78–1.93). The low incidence of cervical cancer (about 6.6 per 100,000 North American women) makes it a challenge to assess its relation to SLE with sufficient power. There may also be ascertainment issues, because cancer registries often do not record noninvasive malignancies.

Nevertheless, the suggested increase in incidence in cervical cancer in SLE is particularly interesting from a pathophysiologic standpoint. There is growing evidence for increased rates of HPV (the infection responsible for cervical cancer), including its high-risk aggressive variants, in women with SLE. Moreover, exposure to immunosuppressant medications (especially cyclophosphamide, and possibly azathioprine) in SLE has also been shown to increase susceptibility to high-risk HPV infection, overall HPV infection, and cervical dysplasia. Thus, although a direct link has yet to be found, based on this evidence and the strong association that has been repeatedly shown between cervical HPV infection and malignant transformation, female patients with SLE are likely to be at increased risk for cervical cancer.

As with cervical carcinoma, the risk of other malignancies that are also strongly associated with HPV is increased in SLE. The most up-to-date analyses argue for an increased risk in vulvar carcinoma (SIR, 3.78; 95% CI, 1.52–7.78) with strong trends for an increased risk also for vaginal (SIR, 3.80; 95% CI, 0.46–13.74) as well as anal carcinoma (SIR, 26.9; 95% CI, 8.7–83.4). Although these results are supported by other investigators (Dreyer and colleagues ), the low absolute number of these rare malignancies should be kept in mind. It has been suggested by some clinicians that the increased incidence of these cancer types in SLE could also be accounted for by inadequate viral clearance. Iatrogenic immunosuppression and genetic and innate immunity defects in lupus are possible mechanisms.

Routine screening for cervical dysplasia is important for patients with SLE. However, one study observed that female patients with SLE with the most severe disease burden (based on SLICC/American College of Rheumatology damage index scores) were the least likely to have undergone cervical screening. The explanation is likely multifactorial but it argues for the importance of counseling adherence to national cervical screening protocols and guidelines.