In this review, the available data regarding the risk of lymphoma, skin cancers, and other malignancies associated with biological agents that are approved and those under investigation for use in inflammatory bowel disease (IBD) are highlighted. How providers may approach the use of these agents in various clinical scenarios is discussed. This review may help providers better understand the true risk of malignancy associated with these agents, thereby leading to an enhanced communication process with patients with IBD when therapeutic decisions are being made.
Key points
- •
The risk of lymphoma and skin cancer with anti–tumor necrosis factor (anti-TNF) therapy is largely driven by the use of concomitant immunomodulator therapy, particularly thiopurines.
- •
The risk of lymphoma and skin cancer with newer biologics seems to be similar to that seen with anti-TNF therapy, but this is largely based on either non–inflammatory bowel disease (IBD) data or IBD data with short follow-up.
- •
Tofacitinib has a measurable risk of malignancy when used in patients with rheumatoid arthritis, and further long-term data are needed in patients with IBD.
- •
In patients at particular risk for developing a malignancy while on therapy, there should be an effort to focus on modifiable risk factors.
- •
The optimal approach to the use of biologics in patients with IBD with a previous history of malignancy remains unclear.
Introduction
The pathogenesis of inflammatory bowel disease (IBD) is multifactorial and involves a complex interaction between genetic, environmental, microbial, and immune factors. The ensuing pathologic response affects both the innate and adaptive immune systems, with the net result of these cellular mechanisms being granulocyte accumulation, cytokine production, and intestinal inflammation. This complex immune response has provided the opportunity for the development of biological agents that target specific components of the immune cascade. These biological drugs have dramatically changed the manner in which the treatment of IBD is approached and the outcomes that may be expected. Despite the proven efficacy of these agents and the shifting of treatment goals toward early aggressive use of these drugs, concerns surrounding toxicity have posed a significant emotional barrier to their use. Infectious complications remain the most common side effect, but the fear of developing cancer poses the greatest obstacle for patients and providers to overcome.
Patients with IBD are at an increased baseline risk for intestinal and extra-intestinal malignancies, but the 2 cancers that have gained the most attention are lymphoma and skin cancer. Although there are some conflicting data, several studies suggest that IBD alone is not associated with an increased risk for lymphoma. In contrast, there seems to be a clear association between IBD and the development of melanoma and nonmelanoma skin cancers (NMSC). There is also much evidence implicating IBD-related medications, particularly thiopurines, in the development of these malignancies. It is difficult to determine the risk of malignancy with biologics on their own, because most patients with IBD treated with biologics have also previously been exposed to thiopurines.
In this review, the available evidence regarding the risk of malignancy when using biological agents for the treatment of IBD are highlighted, with a particular focus on lymphoma and skin cancer. This risk is compared with that seen with other treatment options in IBD and with the general population. The approach to managing these medications in patients who have a history of malignancy is further highlighted. We anticipate that this review will help providers better understand how best to broadly approach the difficult conversation of biological therapy and cancer with their patients.
Introduction
The pathogenesis of inflammatory bowel disease (IBD) is multifactorial and involves a complex interaction between genetic, environmental, microbial, and immune factors. The ensuing pathologic response affects both the innate and adaptive immune systems, with the net result of these cellular mechanisms being granulocyte accumulation, cytokine production, and intestinal inflammation. This complex immune response has provided the opportunity for the development of biological agents that target specific components of the immune cascade. These biological drugs have dramatically changed the manner in which the treatment of IBD is approached and the outcomes that may be expected. Despite the proven efficacy of these agents and the shifting of treatment goals toward early aggressive use of these drugs, concerns surrounding toxicity have posed a significant emotional barrier to their use. Infectious complications remain the most common side effect, but the fear of developing cancer poses the greatest obstacle for patients and providers to overcome.
Patients with IBD are at an increased baseline risk for intestinal and extra-intestinal malignancies, but the 2 cancers that have gained the most attention are lymphoma and skin cancer. Although there are some conflicting data, several studies suggest that IBD alone is not associated with an increased risk for lymphoma. In contrast, there seems to be a clear association between IBD and the development of melanoma and nonmelanoma skin cancers (NMSC). There is also much evidence implicating IBD-related medications, particularly thiopurines, in the development of these malignancies. It is difficult to determine the risk of malignancy with biologics on their own, because most patients with IBD treated with biologics have also previously been exposed to thiopurines.
In this review, the available evidence regarding the risk of malignancy when using biological agents for the treatment of IBD are highlighted, with a particular focus on lymphoma and skin cancer. This risk is compared with that seen with other treatment options in IBD and with the general population. The approach to managing these medications in patients who have a history of malignancy is further highlighted. We anticipate that this review will help providers better understand how best to broadly approach the difficult conversation of biological therapy and cancer with their patients.
Anti–tumor necrosis factor therapy
Risk of Lymphoma
The hallmark of active IBD is infiltration of the lamina propria by innate (neutrophils, macrophages, dendritic, and natural killer T cells) and adaptive immune cells (B and T cells). The increased number of these activated cells in the intestinal mucosa leads to enhanced local levels of proinflammatory cytokines, which play important roles in the interaction between immune and nonimmune cells. One of these cytokines, tumor necrosis factor α (TNF-α), has been shown to play a key role in the pathogenesis of IBD, and has therefore served as the prototypical target for biologics in both Crohn disease (CD) and ulcerative colitis (UC). Around the time that infliximab was approved from use in CD, there was already increasing concern over a potential increase in risk of lymphoma with these agents. Since then, the evidence surrounding the potential association between anti-TNF therapy and lymphoma occurrence has increased significantly.
Initial attempts at understanding the risk of lymphoma with anti-TNF therapy used pooled analyses of randomized controlled trials (RCTs). The first meta-analysis to look at this subject pooled data from 21 RCTs and assessed the efficacy and safety of anti-TNF agents for the treatment of CD. This study examined 3341 patients exposed to anti-TNF with a median follow-up period of 24 weeks per patient and did not find an increased risk for lymphoma occurrence with anti-TNF use. In most of these RCTs, many of the controls also had some exposure to anti-TNF therapy. The investigators performed a sensitivity analysis of studies with unexposed controls, but this was an exploratory subanalysis and was likely underpowered to address this outcome. A more recent meta-analysis looking specifically at the risk of lymphoma with adalimumab therapy included 6 RCTs with 1594 patients with CD and 3050 patient-years (PY) of exposure (PYE) to adalimumab. These investigators reported that patients treated with anti-TNF therapy in combination with an immunomodulator had a significantly higher risk of lymphoma occurrence when compared with the general population ( Table 1 ). This increase in lymphoma risk with anti-TNF combination therapy had previously been reported in a meta-analysis by Siegel and colleagues. This group pooled data from 26 studies, including 8905 patients with 21,178 patient years of follow-up (PYF), and reported that the risk of lymphoma with anti-TNF combination therapy was increased when compared with the general population (standardized incidence ratio [SIR], 3.23; 95% confidence interval [CI], 1.5–6.9), but when compared with immunomodulator monotherapy in CD, the use of anti-TNF combination therapy seemed to have no incremental increase in risk of lymphoma occurrence (SIR, 1.7; 95% CI, 0.5–7.1). When examining the pediatric gastroenterology literature, a recent systematic review by our group studied 5528 patients with 9516 PYF and reported that the risk of lymphoma with anti-TNF therapy in pediatric IBD was similar to that seen in the general pediatric population (SIR, 3.5; 95% CI, 0.35–19.6), a population of pediatric patients receiving thiopurine monotherapy (SIR, 0.47; 95% CI, 0.03–6.44), and among adult patients with IBD treated with anti-TNF agents (SIR, 0.34; 95% CI, 0.04–1.51).
| Incidence Rate (per 10,000 PYF) | SIR | 95% CI | |
|---|---|---|---|
| Current Thiopurine Without Anti-TNF Exposure | |||
| Herrinton et al, 2011 | 4.1 | 1.4 | 1.2–1.7 |
| CESAME , a | 8.8 | 6.5 | 3.5–11.2 |
| Khan et al, 2013 | 14.6 | 7.5 | 4.7–12.0 |
| Current Thiopurine with Previous Anti-TNF Exposure | |||
| CESAME , a | 8.8 | 6.5 | 3.5–11.2 |
| Herrinton et al, 2011 | 15.1 | 5.3 | 3.5–7.0 |
| Current Anti-TNF Therapy with Current Thiopurine Therapy | |||
| Dulai et al, 2014 | 2.1 | 3.5 | 0.35–19.6 |
| TREAT | 4.5 | 2.0 | 0.87–3.95 |
| Siegel et al, 2009 | 6.1 | 3.2 | 1.5–6.9 |
| CESAME | 10.4 | 10.2 | 1.2–36.9 |
| Osterman et al, 2014 | 14.3 | 8.0 | 0.97–29.0 |
| Herrinton et al, 2011 | 19.1 | 6.6 | 4.4–8.8 |
| Current Anti-TNF Therapy with Previous Thiopurine Exposure | |||
| Herrinton et al, 2011 | 14.9 | 5.2 | 3.5–6.8 |
| Current Anti-TNF Therapy Without Thiopurine Exposure | |||
| n/a | — | — | — |
a Analysis included patients who had who had never received anti-TNF therapy and those who had discontinued anti-TNF therapy.
Although these RCT data seem reassuring, data from observational studies are conflicting. The claims-based population study by Marehbian and colleagues reported that the odds of developing a lymphoma was numerically, but not statistically, higher with anti-TNF monotherapy (odds ratio [OR], 2.1; 95% CI, 0.82–5.40) than with immunomodulator monotherapy (OR, 1.66; 95% CI, 0.95–2.88). A later population-based study using the Kaiser Permanente IBD Registry evaluated lymphoma risk among 16,023 patients with IBD followed for an average of 5.8 years. This study similarly found that when compared with the general non-IBD Kaiser population, the risk of lymphoma with anti-TNF therapy (SIR, 4.4; 95% CI, 3.4–5.4) was higher than the risk of lymphoma with thiopurine monotherapy (SIR, 1.4; 95% CI, 1.2–1.7). These studies therefore suggest that in contrast to RCT, the use of anti-TNF therapy outside clinical trials increases the risk of lymphoma above that seen with immunomodulators. This theory may be explained by the strict selection of patients and exclusion of high-risk populations (elderly) in clinical trials, but there are several key points to consider when interpreting these data. The study by Marehbian and colleagues used a claims-based method for data collection, and therefore is at risk for underreporting or overreporting certain outcomes based on claims submissions and follow-up. Furthermore, the use of pharmacy-based claims is limited by the lack of confirmation regarding medication usage and inability to accurately quantify the residual effects of treatments after discontinuation. These factors may help account for the fact that this study is the only one to have ever reported a lower odds of lymphoma with anti-TNF and immunomodulator combination therapy when compared with monotherapy for either agent (OR, 0.63; 95% CI, 0.08–4.64). Although the Kaiser study by Herrinton and colleagues also noted that anti-TNF therapy was associated with an increased risk of lymphoma, the comparison with control individuals who did not have IBD and use of pharmacy-based confirmation for medication usage similarly limits the conclusions that can be drawn from these data. Given the inherit limitations within RCT and claims-based population studies, perhaps the most compelling evidence comes from prospective observational cohort studies specifically aimed at addressing the question of malignancy risk with IBD treatments.
We have 2 such studies to consider: the Cancers Et Surrisque Associé aux Maladies inflammatoires intestinales En France (CESAME) trial and the Crohn’s Therapy, Resource, Evaluation, and Assessment Tool (TREAT) registry. The TREAT registry enrolled 6273 patients with an average follow-up of 5.2 years/patient and found that neither immunomodulator monotherapy (hazard ratio [HR], 1.43; 95% CI, 0.92–2.21), nor infliximab monotherapy (HR, 0.59; 95% CI, 0.28–1.22), nor their combination (HR, 1.22; 95% CI, 0.81–1.86) was independently associated with an increased risk of malignancy. When looking specifically at the risk of lymphoma, there appeared to be no statistical difference in incidence between patients treated with infliximab (SIR, 2.01; 95% CI, 0.87–3.95) or immunosuppressants (SIR, 2.06; 95% CI, 0.99–3.80), when compared with the general population. In contrast, the CESAME trial reported that patients continuing thiopurine therapy in combination with anti-TNF therapy (SIR, 10.2; 95% CI, 1.24–36.9) were at a higher risk for lymphoma when compared with patients continuing thiopurine therapy without anti-TNF therapy (SIR, 6.5; 95% CI, 3.48–11.2).
Hepatosplenic T-Cell Lymphoma
A subtype of lymphoma, hepatosplenic T-cell lymphoma (HSTCL), has raised the greatest amount of concern when using biologics, given the aggressiveness and fatality rate of this malignancy. A study to address this concern pooled data from published literature and the US Food and Drug Administration (FDA) Adverse Event Registry System (AERS). At the time of their publication in 2011, there had been 36 cases reported, with all but 2 occurring in male patients (12–58 years of age). All 36 patients had received thiopurine therapy alone (n = 16) or in combination with an anti-TNF agent (n = 20). A subsequent review of the FDA AERS 2 years later identified only 1 additional case in a patient with CD. This study used 3 causality assessment tools and noted that establishing a causative effect between anti-TNF therapy and HSTCL beyond possible was not feasible. A more recent study using AERS and a literature search examined the risk of subtypes of all T-cell lymphomas across different disease entities (IBD n = 45, rheumatoid arthritis [RA] n = 38, psoriasis n = 11, ankylosing spondylitis n = 6). These investigators found that the risk of T-cell lymphoma was highest in patients receiving thiopurine monotherapy and combination therapy with an anti-TNF agent and thiopurines. A single case of HSTCL was reported with anti-TNF monotherapy, which was not statistically significantly increased when compared with a control population. Although these studies give us the best information currently available for HSTCL, they all use the AERS system, which has inherent biases, which are difficult to overcome. More information is still needed to better understand the absolute incidence of these nearly universally fatal malignancies and how much of an influence anti-TNFs play (or do not play) in their genesis. Similar to the standard lymphomas reported earlier, it is likely that most, if not all, of the risk is associated with thiopurine therapy.
Risk of Skin Cancer
NMSC
NMSC includes both basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), and this cancer ranks among the most common malignancies reported in the United States, with an annual incidence of more than 3 million cases. Although most NMSC have a low metastatic potential and can be easily treated with topical and surgical therapies, those occurring in high-risk areas may have up to a 30% risk of metastasis, and the treatment of these lesions may be costly and lead to permanent disfiguration. Given the increased baseline risk of NMSC in patients with IBD, and the impact that this malignancy has on patient outcomes, it is important to understand how IBD therapy may affect this risk.
The first study to directly address this question was a nested case-control retrospective cohort analysis in which 742 patients with IBD with NMSC were matched to 4 patients with IBD without NMSC to understand the impact of immunomodulators and biologics on malignancy risk. This group reported that the recent (<90 days) and persistent (>365 days) use of biological monotherapy increased the risk of NMSC 2-fold to 3-fold, but this risk was still lower than the 3-fold to 4-fold increase in risk seen with thiopurine monotherapy and the 5-fold to 6-fold increase in risk seen with combination therapy ( Table 2 ). However, a later study by the same group using similar methodology found the risk of NMSC with biologics to be limited to only those patients with persistent (>365 days) use of biologics. These investigators again reported that the use of an immunomodulator, alone or in combination with a biologic, carried a higher risk for NMSC than biological therapy alone. When comparing this risk with the general population, a recent meta-analysis pooled data across all adalimumab RCTs and found that the use of adalimumab monotherapy was not associated with an increased risk of NMSC (SIR, 1.20; 95% CI, 0.39–2.80), whereas the use of adalimumab combination therapy with an immunomodulator was associated with an increased risk of NMSC when compared with the general population (SIR, 4.59; 95% CI, 2.51–7.70).
| Treatment | OR | 95% CI | |
|---|---|---|---|
| Long et al, 2010 | Recent use (<90 d) | ||
| Biologics | 2.47 | 1.29–4.73 | |
| Immunomodulators | 3.71 | 2.74–5.02 | |
| Combination therapy | 5.85 | 3.2–10.8 | |
| Persistent use (>365 d) | |||
| Biologics | 3.23 | 1.24–8.45 | |
| Immunomodulators | 4.45 | 2.94–6.75 | |
| Combination therapy | 6.75 | 2.74–16.65 | |
| Long et al, 2012 | Overall risk | ||
| Biologics | 1.14 | 0.95–1.36 | |
| Immunomodulators | 1.85 | 1.66–2.05 | |
| Persistent use (>365 d) | |||
| Biologics | 1.63 | 1.12–2.36 | |
| Immunomodulators | 2.72 | 2.27–3.26 | |
| Combination therapy | 3.89 | 2.33–6.46 | |
| Osterman et al, 2014 | Combination therapy | 3.46 | 1.08–11.06 |
Melanoma
In contrast to NMSC, melanomas are not so easily cured, and the mortality secondary to these cancers has been largely unchanged in the United States despite efforts to enhance early detection of these lesions. Patients with IBD are at a higher baseline risk for the development of melanoma skin cancer, but when attempting to understand the added risk of melanoma with anti-TNF therapy, there are few data on which to base decisions. The nested case-control study by Long and colleagues reported that, unlike NMSC, the risk of melanoma was increased with biologics (OR, 1.88; 95% CI, 1.08–3.29) but not immunomodulators (OR 1.10; 95% CI, 0.72–1.67). In an effort to understand this association, the investigators performed an exploratory subgroup analysis and identified that patients undergoing long-term therapy with anti-TNF agents were at higher risk than those not exposed to long-term therapy (OR, 3.93; 95% CI, 1.82–8.50). In contrast to this study, the CESAME group reported that the risk of melanoma was not increased in patients with IBD who were receiving thiopurines (SIR, 1.09; 95% CI, 0.13–3.94) or anti-TNF treatment (SIR, 2.57; 95% CI, 0.06–14.33). Less than 10% of patients in the CESAME trial were prescribed nonthiopurine medications, and the investigators therefore commented that a reliable risk of melanoma with anti-TNF therapy could not be determined.
Other biologics
Although anti-TNF agents now represent an integral component of the treatment algorithm for IBD, less than a third of patients initially responding to these medications maintain remission beyond 1 year of therapy. Researchers have therefore also turned their attention toward several other therapeutic targets, including proinflammatory cytokines (interleukin 12 [IL-12]/IL-23), the downstream effects of cytokine stimulation (leukocyte trafficking), and the cytokine signaling pathways (Janus kinase [JAK]) responsible for the altered immune response and inflammation seen in IBD. These newer biologics represent a new article in the treatment paradigm of IBD, and it is therefore imperative to understand any associated risk of malignancy with these agents to better optimize the long-term management of patients, particularly given the role that these drug targets have in immune function and neoplastic transformation. Data regarding the risk of malignancy with these agents are largely limited to non-IBD populations or early phase RCTs used for drug approval in patients with IBD.
Anti-Integrins (Natalizumab and Vedolizumab)
Natalizumab is a recombinant humanized monoclonal IgG4 antibody directed against the integrin subunit α 4 , which blocks both α 4 β 7 receptors in the gut and α 4 β 1 receptors in the central nervous system (CNS). Natalizumab was the first antiadhesion treatment developed and approved for use in IBD, but, despite its proven efficacy, its use has now largely been limited by the risk of developing progressive multifocal leukoencephalopathy, an opportunistic and usually fatal brain infection caused by reactivation of latent John Cunningham polyomavirus. Long-term follow-up data are largely lacking in the population with IBD, but reports from patients who do not have IBD have raised concern that therapies targeting cell adhesion molecules may be associated with an increased risk of cancer progression. Initial reports noted a possible association between natalizumab therapy and the development of primary CNS lymphoma in patients with multiple sclerosis, but when looking at pooled cumulative clinical trial experience, there appeared to be no clear association, and the rate of lymphoma with this therapy was quantified to be lower than that seen with other biologics (1.3/10,000 PYF). Similarly, case reports suggested that natalizumab may be associated with an increased risk for melanoma, but this was not supported by a later meta-analysis of clinical trials, which reported that the rate of melanoma with natalizumab (4.2/10,000 PYF) was similar to that seen with placebo (8.2/10,000 PYF).
Vedolizumab is a humanized IgG1 monoclonal antibody that binds exclusively to the gut-specific α 4 β 7 -integrin and inhibits leukocyte trafficking into the gastrointestinal mucosa. This selective blockade of gut-homing T lymphocytes allows it to avoid the CNS side effects noted with previous anti-integrins (ie, natalizumab). Two recent phase 3 RCTs have reported that vedolizumab was more effective than placebo in inducing and maintaining remission in both CD and UC. An interesting observation within these studies was the differential effect of vedolizumab for inducing a clinical response in CD and UC, with patients with CD requiring a longer duration of therapy to show a significant benefit, which was postulated to be secondary to a time-dependent inhibition of leukocyte trafficking in CD or the more systemic nature of the disease. These data support the use of vedolizumab in both CD and UC and support the long-term use of this agent in IBD, making an assessment of cancer risk important. Given that vedolizumab is not approved for any other indication, these phase 3 RCTs are the only data that determine the risk of malignancy with this agent. These RCTs in combination with ongoing long-term safety studies include a total of 3129 patients with 3096 PYF. Within this cohort, 1 lymphoma (B-cell lymphoma), 3 NMSC (2 SCC, 1 BCC), and 2 melanomas have been reported ( Table 3 ). This single lymphoma occurred in a 42-year-old man with CD who had received previous immunosuppressive therapy, including infliximab and azathioprine, and continued on azathioprine and corticosteroids while enrolled in the study. The patient had worsening CD despite therapy (21 doses of vedolizumab) and underwent colorectal surgery. Approximately 6 weeks later, the patient was found to have a solid mass in the lesser pelvis in the vicinity of the rectal stump, for which he was discontinued from the study, and received treatment with rituximab. Of the 5 skin cancers, only 2 were believed to be possibly related to vedolizumab when reviewed by independent study investigators or an FDA reviewer.
Related posts:
Anti–Tumor Necrosis Factor-α Monotherapy Versus Combination Therapy with an Immunomodulator in IBD
Update on Anti-Tumor Necrosis Factor Agents in Crohn Disease
Pharmacokinetics of Biologics and the Role of Therapeutic Monitoring
Biologics of IBD
An Update on Anti-TNF Agents in Ulcerative Colitis
Update on Janus Kinase Antagonists in Inflammatory Bowel Disease
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
