Article Text

Download PDFPDF

The management of immunosuppression in patients with inflammatory bowel disease and cancer
  1. Oren Bernheim1,
  2. Jean-Frederic Colombel1,
  3. Thomas A Ullman1,
  4. David Laharie2,
  5. Laurent Beaugerie3,
  6. Steven H Itzkowitz1
  1. 1 The Dr Henry D Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, New York, New York, USA
  2. 2 Service d'Hepato-Gastroenterologie, CHU de Bordeaux, Hopital Haut-Leveque, Bordeaux, France
  3. 3 Department of Gastroenterology, AP-HP, Hôpital Saint-Antoine, Paris, France
  1. Correspondence to Professor J-F Colombel, The Dr Henry D Janowitz Division of Gastroenterology, Icahn School of Medicine at Mount Sinai, One Gustave L Levy Place, Box 1069, New York, NY 10029, USA; jean-frederic.colombel{at}

Statistics from


Inflammatory bowel diseases (IBD) are chronic inflammatory conditions of the gastrointestinal tract. Immunomodulators and biologic agents (hereafter referred to jointly as ‘immunosuppression’) are effective in treating IBD and recent evidence supports their introduction earlier in the disease course.1 ,2 An important concern to both patients and physicians considering immunosuppression for the treatment of IBD is the potential associated cancer risk.

It is estimated that 1.6 million new diagnoses of cancer will be made in 2013 among the general population in the USA alone.3 Likewise, as the population of patients with IBD ages, it is inevitable that some will develop cancer. As such, in addition to concerns about the development of de novo cancer in a patient without a history of cancer, clinicians caring for patients with IBD are increasingly challenged with questions about IBD management after the development of cancer. Conversely, oncologists are often concerned about how to manage IBD in cancer patients.

Several important clinical questions deserve further attention with respect to IBD therapy and cancer (figure 1). First, does medical therapy for IBD predispose to developing cancer? Second, in an IBD patient with a history of cancer, does IBD therapy impact cancer recurrence? Third, once cancer develops in an IBD patient, is the cancer outcome different? Finally, in an IBD patient with current cancer, does the cancer therapy affect IBD outcomes?

Figure 1

Clinical questions related to patients with inflammatory bowel disease and cancer. IBD, inflammatory bowel disease; TNF, tumour necrosis factor.

Question 1, dealing with the topic of de novo cancer development related to treatments for IBD in patients without a history of cancer, has recently been reviewed.4 Less is known about the impact of immunosuppressive therapies on cancer in IBD patients who have had a prior (controlled) cancer or have a current (uncontrolled) cancer. The purpose of this article is to review the role of immunosuppression on cancer outcomes and to suggest strategies for the treatment of patients with IBD and a prior or current cancer.

Management of IBD in patients with a prior cancer

Currently, there are no consensus guidelines to assist in managing IBD patients with prior cancer (figure 1, question 2). Most clinical data on the effect of immunosuppression has come from observational studies of patients with rheumatoid arthritis (RA) or solid organ transplants. In fact, in the pursuit of regulatory approval, randomised controlled trials of antitumour necrosis factor (anti-TNF) therapies in patients with IBD purposely excluded patients with active malignancy, or a history of cancer within the last 5 years, over concern for provoking cancer progression or recurrence. Consequently, there are no randomised prospective data in this population.

When considering immunosuppression in patients with prior cancer, two potential scenarios should be considered. A prior cancer may be completely eradicated which should theoretically eliminate the risk of recurrence under immunosuppression. Alternatively, a prior cancer may be dormant due to treatment effects and ongoing immunosurveillance and will therefore have a quantifiable risk of recurrence under immunosuppression. Unfortunately, these two scenarios are typically clinically indistinguishable.

Lessons from the transplant population

The concern about immunosuppression in IBD patients with a current or prior cancer is that the immune modulation may exacerbate the tumour. Lessons from patients with organ transplants are instructive because they are among the most strongly immunosuppressed. In patients with a prior cancer, Penn noted that the cancer recurrence rate was >20% in the years following renal transplantation.5 As a variety of immunosuppressive regimens were used, no correlation could be made between recurrence and any particular immunosuppressive agent. Based on these findings, they assigned a relative risk of cancer recurrence for various malignancies (table 1). A follow-up study including 1137 neoplasms demonstrated that the recurrence risk was highest (54%) when transplantation occurred within 2 years following cancer treatment compared with 33% at 2–5 years and 13% at more than 5 years after cancer treatment.6

Table 1

Risk of cancer recurrence based on pre-existing cancer subtype in patients on immunosuppression following renal transplantation (adapted from Penn5)

In the transplant population, length of exposure and intensity of therapy with immunosuppressants increases the de novo malignancy risk.7 Interestingly, transplant patients treated with mammalian target of rapamycin (mTOR) inhibitor therapy had an overall lower risk of post-transplant de novo malignancy compared with those treated with calcineurin inhibitors.8–10 Indeed, following liver transplantation for hepatocellular carcinoma, a meta-analysis suggested that using the mTOR inhibitor sirolimus instead of a calcineurin inhibitor resulted in decreased tumour recurrence rates and higher recurrence free survivals.11 To date, mTOR inhibitors have been ineffective in Crohn's disease.12

Thiopurine experience

The CESAME cohort followed 405 IBD patients with a prior cancer for 2.9 years.13 Rates of new and recurrent cancer were compared between 93 patients exposed and 312 patients not exposed to a thiopurine. No significant difference was identified in the crude incidence rate of new or recurrent cancer between the two groups. This study included 232 patients with a recent (<5 years from enrolment) cancer diagnosis and may be the most compelling evidence to date regarding the safety of resuming a thiopurine in an IBD patient with a prior cancer.

Anti-TNF experience

To date, data on anti-TNF use in patients with autoimmune diseases and a history of a prior cancer come from studies of RA patients, and not IBD patients. In 2010, Dixon et al, using the British Society for Rheumatology biologics register, showed no significant difference in the risk of cancer recurrence in 177 RA patients with a prior cancer who were later exposed to anti-TNF therapy, compared with 117 patients with a prior cancer treated only with disease modifying antirheumatic drug (DMARD) therapy.14 They observed a lower risk of cancer recurrence in the anti-TNF group than might have been discovered if their malignancy had been more recent. This study was limited by the small sample size and the fact that, compared with the DMARD treated group, the anti-TNF group had a more remote history of malignancy, which might have contributed to their lower rate of recurrence. Of note, a history of melanoma carried the highest risk of recurrence in the anti-TNF treated group (18%) compared with the DMARD group (0%). Data from the German biologics register reported no difference in cancer recurrence rates in RA patients with a history of cancer exposed to anti-TNF (n=55) or traditional DMARD (n=58) therapy.15

Despite large registries of over 10 000 patients exposed to anti-TNF, few exposed patients have an identifiable history of cancer. It may be that providers are hesitant to start anti-TNF agents in such patients due to theoretical or medicolegal concerns. Indeed, at the time of enrolment into the British biologics register, the British Society for Rheumatology guidelines recommended that ‘caution should be exercised in the use of anti-TNF therapies in patients with previous malignancy’—the likely reason for the selection differences noted in the Dixon study.16

To date, IBD management guidelines rarely address the patient with a history of cancer. The 2010 European Crohn's and Colitis Organisation guidelines indicate that anti-TNF therapy is contraindicated in patients with a history of lymphoma and ‘careful consideration should be given to initiating anti-TNF therapy’ in those with a history of non-haematopoietic cancer.17 The 2009 American College of Gastroenterology and 2010 World Gastroenterology Organisation guidelines do not include any recommendations regarding the management of IBD in patients with a history of cancer.18 ,19

While the data presented above in RA patients may be reassuring, prospective data with a larger number of patients, and those with more recent cancers, are needed before a more definitive statement on safety can be made. Also, patients with IBD should be studied to explore the relative safety of using anti-TNF agents in patients with cancer. Even if additional reassuring data emerge, the theoretical concern of contributing to cancer recurrence will have to be overcome by the provider and the patient.

Management of IBD in patients with current cancer

It is generally accepted that when a cancer is diagnosed in patients with IBD who are receiving immunosuppressants, practitioners should agree to a hiatus in the immunosuppressive therapy until the cancer is controlled (figure 1, questions 3 and 4).20 The MICISTA registry, a database of 7158 IBD patients evaluated at St Antoine Hospital from 1975 onwards, found that the diagnosis of cancer led to IBD therapy modification (decreased immunosuppression) in a large proportion of patients.21 Conversely, an observational study from Denmark found that patients with breast cancer who had Crohn's disease were treated less often with radiotherapy and also had a higher mortality compared with non-IBD patients with breast cancer, suggesting that oncologists may avoid certain cancer therapies in patients with IBD, perhaps to their detriment.22 Together, these data suggest that oncologists alter treatment plans for cancer in patients with IBD, and gastroenterologists alter IBD treatment plans in patients with cancer. Ultimately, these modifications may significantly impact on the course of both IBD and cancer, and the consequences of these alterations over the course of either disease are largely unknown.

It is also important to consider how prior exposure to immunosuppression might impact on cancer outcomes in patients with an active malignancy. Observational data in 314 patients with RA and cancer who had been previously exposed to anti-TNF therapy found no significant difference with regard to the stage of cancer at the time of diagnosis, or the prognosis and risk of death compared with 586 matched biologic naïve controls.23 Similarly, a small retrospective study of patients with IBD and lymphoma found no difference in survival rates between those previously exposed (n=7) and those not exposed (n=7) to a thiopurine.24

Thiopurine experience

Thiopurines are believed to have both carcinogenic and anticancer properties.25 Azathioprine and 6-mercaptopurine are metabolised into 6-thioguanine, which incorporates into DNA and RNA ultimately resulting in cytotoxicity. While this pathway results in immunosuppression, it may also be responsible for promoting cancer.26 Also, some patients who develop malignancy while taking thiopurines may have spontaneous remission of the cancer on withdrawal of the thiopurine.27 ,28 Of additional relevance to patients with cancer, thiopurines are known to cause bone marrow suppression. Oncologists may prefer to discontinue a thiopurine at the time of a cancer diagnosis due to the presumed risk of severe bone marrow suppression should cytotoxic chemotherapy be planned to control the cancer.

Anti-TNF experience

TNF, as well as anti-TNF agents, have been studied for their effects on cancer. Despite having a name implying that it will induce necrosis of tumours, the impact of TNF on malignancy is unpredictable.20 On the one hand, TNF can initiate cellular apoptosis by activating the extrinsic pathway, which involves death ligands/receptors that signal through caspase 8 and 10 activation.29 Both intrinsic and extrinsic pathways converge on caspases 3, 6 and 7, whose actions effectuate apoptosis.30 Accordingly, the pro-apoptotic role of TNF has prompted the theoretical concern that inhibiting TNF may cause recurrence or rapid tumour progression in patients with cancer. On the other hand, TNF has also been shown to facilitate survival and proliferation of neoplastic cells via the nuclear factor κB cascade.31 Thus after several studies in the 1990s showed limited benefit of TNF in the treatment of cancer (mainly due to intolerable side effects), interest shifted towards blocking TNF as a potential treatment for solid tumours.31

A phase II study of infliximab (IFX) in 37 patients with immunotherapy resistant or refractory renal cell carcinoma concluded that blocking TNF-α may actually be beneficial.32 Another report on the safety profile and biological response of IFX in 41 patients with various cancers found no evidence of disease acceleration.33 A phase II randomised study of IFX plus gemcitabine to treat cachexia in patients with pancreatic cancer reported no evidence of disease acceleration or increased mortality.34 Additionally, a placebo controlled double blind trial assessing cancer associated weight loss in 61 elderly patients with metastatic non-small cell lung cancer found no statistically significant difference in tumour response/stabilisation or overall survival.35 Although these studies demonstrate the potential safety of administering anti-TNF agents to patients with an active cancer, they should be interpreted with caution, given that the patient population was limited to cases of advanced cancer with an overall poor prognosis. On the other hand, patients with chronic obstructive pulmonary disease who are exposed to IFX may be at a higher risk of developing lung cancer.36 Additionally, there is a case report of a patient who developed non-small cell lung cancer while undergoing adalimumab therapy for Crohn's disease who went into a sustained remission on withdrawal of the anti-TNF.37

Indirect evidence also comes from studies on ipilimumab, a monoclonal antibody directed against cytotoxic T lymphocyte associated antigen 4 that is being increasingly used in patients with advanced melanoma or renal cell carcinoma. This agent itself can cause a drug induced colitis.38 IFX was a highly effective treatment for this severe form of colitis and resulted in prompt improvement when steroids failed.38 ,39 Although only a handful of cases were reported, there were no adverse cancer clinical outcomes from the administration of IFX to treat the ipilimumab associated colitis in these patients but authors advise caution when administering anti-TNF to patients with an active malignancy.

Management of current cancer in patients with IBD

Axelrad et al 40 analysed 84 patients with IBD who had an extraintestinal solid malignancy (figure 1, question 4). Among 69 patients with inactive IBD within 6 months of cancer diagnosis, 17.4% flared during follow-up. The strongest predictor of flare was the type of therapy administered: hormonal therapy (as monotherapy or combined with cytotoxic chemotherapy) was associated with the greatest risk. Of 15 patients with active IBD up to 6 months before the cancer diagnosis, 66.7% achieved IBD remission (five treated with cytotoxic monotherapy, four treated with cytotoxic and hormonal combination therapy, and one treated with hormonal monotherapy). The 33.3% who did not achieve IBD remission had all received hormonal monotherapy. These findings suggest that cytotoxic chemotherapy may help control IBD during cancer therapy.

In cases where IBD symptoms break through the chemotherapy regimen, it is not clear whether corticosteroids or anti-TNF should be used. Oncologists frequently use corticosteroids for their effect on tumour and treatment related symptoms as well as their antitumour effects, but they may also enhance tumour cell resistance to apoptosis and decrease immunosurveillance.41 Whether corticosteroids are carcinogenic by themselves remains controversial because the impact of steroids must be distinguished from the impact of the underlying disease and associated immunosuppressants. Some population based studies have suggested an excess risk of lymphomas and non-melanoma skin cancers associated with prolonged exposure to corticosteroids.42 ,43 Although there are few data regarding the impact of steroids on cancer, the general consensus is that corticosteroids are the safer option although anti-TNFs may be a useful back-up plan. Future research efforts should aim to compare steroids and anti-TNF therapy in the management of patients with uncontrolled cancer and uncontrolled IBD.

Conclusions and recommendations

The data reviewed above are summarised in table 2. Although there are limitations of currently available data, we propose the management guidelines in figure 2. Ultimately, management decisions regarding immunosuppression in IBD patients with a cancer are best made in a multidisciplinary fashion and on a case by case basis, taking into account the natural history of the cancer (organ of origin, stage, histological type and prognosis). This should include the opinion of the oncologist, the time from completion of cancer treatment, the severity of the IBD and the expected impact of the immunosuppressant on the previous cancer.

Table 2

Available studies on patients with a prior or current cancer and exposure to immunosuppression

Figure 2

General recommendations for the management of immunosuppression in patients with IBD and a prior cancer or a current cancer. IBD, inflammatory bowel disease; IS, immunosuppression; TNF, tumour necrosis factor.

With the exception of long term adjuvant hormone therapy for cancer, it seems reasonable to withhold immunosuppressants during the treatment of active cancer. Withholding thiopurines may be particularly worthwhile if bone marrow suppression is expected from the planned chemotherapy. Within 2 years after completing cancer treatment, 5-aminosalicylates, corticosteroids, antibiotics, nutritional therapy and surgery should form the foundation of IBD treatment. However, if these treatment modalities are ineffective in a patient with severe IBD, then immunosuppressive therapy should be considered as rescue therapy, despite any potential deleterious effects on cancer risk. If a tumour has an intermediate to high risk of recurrence in the post-transplant state (table 1), an interval of 5 years is preferable. In all cases, if immunosuppressive therapy is resumed, a prudent step-up approach should be considered, starting with monotherapies (perhaps preferring methotrexate, if appropriate) as a first choice.

Many clinical questions remain regarding the treatment of IBD in a patient with cancer. While multicentre prospective registries will continue to be of great importance in studying this patient population, the data analysed from these registries will need to be interpreted with caution, as practitioners have likely selected ‘low risk’ patients with an active or a prior cancer to receive immunosuppression. Ultimately, this entire area calls for more research as more and more IBD patients will be diagnosed with cancers due to their overall increased lifespan. Indeed, as we are improving medical and surgical therapy for IBD as well as enhancing infectious and cardiovascular disease management, we must be ready to manage IBD therapy in an aging population with its medical comorbidities, including cancer.


The authors appreciate the input from the NYC IBD Consortium, New York, NY, USA (chairs: Seymour Katz, Mark Pochapin, Ellen Scherl, Steven Itzkowitz).


View Abstract


  • Contributors J-FC, SHI, TAU and OB were responsible for the concept and design of the paper. All authors contributed to the drafting, writing and critical revisions of the paper. All authors have agreed to the final version to be published.

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.