Article Text
Statistics from Altmetric.com
- Cancer risk
- IBD
- IBD therapy
- cancer risk
- thiopurines
- CESAME
- lymphomas
- colorectal cancer
- 13C-urea breath test
- intestinal bacteria
- IBD clinical
- 2,4,6-trinitrobenzene sulphonic acid
- inflammatory diseases
Although there has been a recent trend towards a decreased absolute age-standardised mortality from cancer in Europe,1 in 2004, for the first time, cancer became the main cause of death in the general population in France.2 This apparent paradox is explained by a faster improvement of the prognosis of non-cancer diseases. Cancer risk has therefore become a lifelong major fear for gastroenterologists and patients when considering the long-term impact of the natural history and treatment of inflammatory bowel diseases (IBDs).
Overall mortality from cancer in IBD
Life expectancy is normal for patients with ulcerative colitis (UC), with no excess mortality from cancer in population-based studies.3 For patients with Crohn's disease (CD), life expectancy is slightly reduced.4 A significant increase in deaths from cancer was reported in the most recent meta-analysis of population-based studies,4 but this was probably due to an over-representation of smoking-related causes of death in a disease where smokers are over-represented. In a Californian medico-administrative database, no overall excess mortality from cancer was demonstrated in either type of IBD, irrespective of the treatments used.5 This reassuring general background must be highlighted before discussion of the specific details of inflammation and immunosuppression-related cancers in IBD.
Theoretical links between IBD and cancer
There is no evidence for a common genetic background between IBD and a susceptibility to cancer. For environmental factors, increased and decreased proportions of smokers in CD and UC compared with the general population are associated with corresponding variations in the incidence of lung and urinary bladder cancers.6 Finally, immunosuppression and inflammation are the two main drivers of IBD-related carcinogenesis. The difficulty is that these two mechanisms may be interlinked, particularly in the intestinal tissues; thus, we will now distinguish the true (negative and positive) effects of IBD treatments.
Cancers related to immunosuppressive therapy
We have learnt from experience with transplantation that immunosuppressive therapy can facilitate, in a dose- and time-dependent manner, the genesis of de novo malignancies, and, once the cancer has developed, accelerate tumour growth and metastasis.7 The mechanisms behind this process are decreased immunosurveillance of emerging cancers,8 facilitation of the action of oncogenic viruses and, for certain immunosuppressive drugs, a direct oncogenic effect.7
Lymphoproliferative disorders
In contrast to other inflammatory chronic diseases, such as rheumatoid arthritis, it seems that the chronic ‘systemic’ inflammation of IBD is not associated with an intrinsic excess risk of extra-intestinal lymphomas. Thus it is found that patients who are not exposed to immunosuppressants in nationwide cohorts have the same age and sex-standardised incidence rates of lymphomas as the general population.9 10
Risk associated with the use of thiopurines alone or in combination with anti-tumour necrosing factor (TNF) therapy
Patients with IBD exposed to thiopurines have a three- to fivefold increased risk of developing lymphoproliferations.9 11 Most of the thiopurine-associated lymphomas are due to the reactivation of a latent Epstein–Barr virus (EBV) infection,12 as is the case with late-onset lymphomas which occur after transplantation.13 Transplant specialists have made considerable progress in avoiding fatal forms of EBV-related lymphoproliferations. Tight monitoring of systemic viral load leads to the detection of EBV-associated lymphoproliferations, particularly in the early post-transplant phase. In this period, patients at a maximum risk are the EBV-seronegative recipients who become infected by the transplant organ from an EBV-positive donor.14 Various forms of pre-emptive or curative treatments (antiviral therapy, rituximab, infusion of EBV-specific T cells, conventional chemotherapy) are being developed.15–17 None of these detection or curative methods has been validated or transposed for use in IBD: these methods are only avenues of research that cannot be recommended in routine practice (table 1). However, patients with controlled IBD receiving immunosuppressants who develop an unexplained fever, lymphadenopathy or haemophagocytic syndrome18 are at risk of an EBV-related early polyclonal lymphoproliferation secondary to EBV reactivation or a late primary EBV infection complicated by lymphoproliferation. Measurement of the systemic EBV viral load should be performed in such cases to optimise the diagnosis.
Young (<35-year-old) men receiving thiopurine treatment who are seronegative for EBV are at risk of developing fatal forms of primary EBV infection, including polyclonal post-mononucleosis lymphoproliferations, such as X-linked lymphoproliferations due to XIAP mutations.19 Provided that such cases of lymphoma have not been reported with anti-TNF therapy, then the use of anti-TNF therapy alone in young EBV-negative men with IBD requiring immunosuppressants is theoretically a way to reduce the risk. Future research should determine whether patients who develop this particular complication have a genetic X-linked background, which might encourage genetic screening before the initiation of thiopurine treatment. Hepatosplenic T cell lymphomas primarily affect young (<35-year-old) men receiving combination therapy with thiopurines and anti-TNF for >2 years.20 Although the absolute risk seems low, limiting the duration of combination therapy to 2 years—that is, in many cases reducing the immunosuppressive therapy as soon as histological and clinical remission is obtained, might limit the problem in patients at risk.
Patients with IBD have a significant excess risk of intestinal lymphoma arising in chronically inflamed intestinal segments.21 Most of these lymphomas are EBV-associated, and epidemiological and pathological studies have suggested a specific pathogenesis,22 23 with a central role for EBV local intestinal infection in inflammation or drug-induced immunosuppression. It is not easy to determine the role of immunosuppressants. The absolute risk of intestinal lymphomas in IBD is low, but when such lymphomas arise within inflammatory lesions, diagnosis may be difficult and delayed. As for other inflammation-related cancers, early and sustained control of IBD lesions with aggressive treatment may reduce the excess risk of intestinal lymphomas in IBD.
Risk associated with the use of corticosteroids, methotrexate and anti-TNF
Corticosteroids
Corticosteroids are used in oncological practice because of their pro-apoptotic effects on lymphoid cells and their beneficial action on some tumour complications, but it has been shown that corticosteroids may also enhance tumour cell resistance to apoptosis and decrease immunosurveillance.24 A high incidence of lymphomas at a population-based level has been reported in regular corticosteroid users.25 26 The long-term use of corticosteroids is discouraged in IBD treatment and is not the only possible treatment and so is now rarely used. Thus the carcinogenic effect of the long-term use of corticosteroids in IBD is not a major problem.
Methotrexate
Methotrexate and anti-TNF are not pure immunosuppressants, are not used after transplantation and are not extensively used as monotherapies in IBD. Our knowledge, therefore, is derived from experience with rheumatological diseases. Whether exposure to methotrexate induces an excess risk of cancer in rheumatological diseases is still debated. As an illustration, methotrexate can, like thiopurines, induce sporadic reversible forms of polyclonal EBV-associated lymphoproliferation in patients with rheumatoid arthritis,27 but, at a nationwide population level, there was no apparent increase in non-Hodgkin's lymphomas in French patients with rheumatoid arthritis who were treated with methotrexate.28 Given the current limited use of methotrexate for long-term monotherapy in IBD, no information on the cancer-specific safety of methotrexate in IBD is likely to be available in the near future.
Anti-TNF alone
TNF has a dual effect on tumour progression through different molecular pathways that are predominantly triggered after binding to the cellular receptor.29 On the one hand, TNF can stimulate apoptosis via the caspase pathway and the necrosis of tumours. On the other hand, TNF can facilitate the survival and proliferation of neoplastic cells via the NF-κB cascade. The clinical impact of anti-TNF therapy on cancer risks in chronic inflammatory diseases is unpredictable. A meta-analysis of randomised controlled trials of infliximab and adalimumab in rheumatoid arthritis concluded that patients exposed to anti-TNF had a threefold increase in the risk of cancer diagnosis, with eight lymphomas reported during the trial in the 3493 patients treated with anti-TNF versus none in the 1512 patients receiving placebo.30 By contrast, in a recent systematic review of registries and prospective observational studies of patients with rheumatoid arthritis, the pooled estimate for the risk of lymphoma in patients exposed to anti-TNF (vs non-exposed) was not significantly increased (RR=1.11, 95% CI 0.70 to 1.51).31
In IBD, the challenges are to determine first, if the spectrum of promoted cancers is the same for anti-TNF as for thiopurines, and second, the overall cancer risk associated with monotherapy with anti-TNF compared with thiopurines alone and with the combination of both treatments. The problem is unresolved because most patients with IBD receive combination therapy. For lymphomas, it was hypothesised in a meta-analysis that the relative risk attributable to anti-TNF was 1.7; this figure was reached by subtracting the risk associated with thiopurine monotherapy from that found in patients receiving combination therapy.32 It must be emphasised that this is pure speculation; a large safety cohort is urgently needed to compare the excess risk of cancers (especially lymphomas and non-melanoma skin cancers (NMSC), but including all other types of cancers) and severe infections observed with monotherapy or combination therapy, with an adjustment for the clinical activity of IBD. If the differences are clinically relevant, this will clearly have an impact on the preferred long-term therapeutic strategies in IBD.
Non-melanoma skin cancers
Risk associated with the use of thiopurines alone or in combination with anti-TNF
The links between NMSC, thiopurines and IBD are now quite clear from the results of three recent studies.33–35 Current exposure to thiopurines is associated with a four- to sixfold-increase in the risk of NMSC. Both subtypes of NMSC are increased: basal-cell skin cancers (BSC), largely predominant in the general population, and squamous-cell cancers (SCC), associated with a poorer prognosis and a greater carcinogenic impact of immunosuppression.36 As a consequence of this latter feature of SCC, the proportion of these cancers in patients with IBD developing NMSC is increased with thiopurine therapy.34 It has been established that thiopurines increase the ultraviolet A radiation-induced DNA damage to skin cells through generation of reactive oxygen species.37 38 Thiopurines may also cause mutations of PTCH, a candidate tumour suppressor gene, which is altered in basal-cell carcinomas. Of note, this latter mechanism is independent of sun exposure.39
In contrast to the promoting effect of thiopurines on lymphomas, which is reversible as soon as the drugs are stopped and the immune control of EBV infection is restored, thiopurine-induced mutations of skin cells seem to have long-term deleterious consequences. Thus patients from the CESAME cohort previously exposed to thiopurines had excess risks similar to those of patients currently exposed.34 This point is crucial, because it supports the recommendation for lifelong sun protection and dermatological screening in all patients with IBD who have ever been treated with thiopurines beyond the first 3 months, during which treatment is often abandoned because of intolerance. Another consequence is that the excess risk of NMSC associated with IBD should be assessed in patients with IBD who were never exposed to thiopurines. In this subgroup of the CESAME population, the risk of NMSC was statistically similar to that of the background general population, suggesting a normal background incidence of NMSC in patients with IBD.34 This also provides a plausible a posteriori explanation for the finding of an apparently ‘IBD-related’ excess risk of BSC in recent studies which did not distinguish between thiopurine users and non-users,6 and in a subgroup analysis with no exhaustive information on thiopurine use in all patients from the diagnosis of IBD.35
Risk associated with the use of anti-TNF alone
In observational cohorts of patients with rheumatoid arthritis, exposure to anti-TNF is associated with a significant excess risk of NMSC (RR=1.45, 95% CI 1.15 to 1.76).31 In IBD, the risk of NMSC associated with the use of anti-TNF alone has not yet been properly assessed. In the cohort-nested case–control study conducted by Long et al, there was a twofold excess risk in patients with IBD exposed to anti-TNF; however, data are not available for exposure to anti-TNF alone.33 In addition, a persistent carcinogenic effect of thiopurines34 may have an impact on the risk of NMSC in patients treated with anti-TNF alone but who were previously exposed to thiopurines.
Uterine cervix abnormalities
The risk of human papillomavirus (HPV)-related cervical abnormalities is increased in HIV-infected or transplant recipient women.7 Whether immunosuppressive therapy alone is an independent risk factor for HPV-related abnormalities in women with IBD among other factors (eg, smoking, oral contraceptives, promiscuity) has not been demonstrated.40–42 However, it seems reasonable to recommend annual screening for cervical abnormalities in women with IBD who are exposed to immunosuppressants, as for immunocompromised individuals.43 Another commonsense measure is to follow the guidelines for HPV vaccine in eligible women,43 particularly before the initiation of the immunosuppressive therapy.44 More problematic is the question of stopping immunosuppressive therapy when cervical abnormalities occur. Drug withdrawal appears to be mandatory when progression or early recurrence of cervical abnormalities occur despite proper management. In other cases, tailored individual decisions in consultation with the gynaecologist are appropriate, balancing all relevant elements (other risk factors for cervical abnormalities, degree of cervical abnormality, HPV subtype if known, need for immunosuppressive therapy and current or expected effect on IBD progression). There is a link between HPV infection and immunosuppressive therapy for some cases of anal carcinoma arising in chronically inflamed anal lesions in Crohn's disease, but the epidemiological proof for a link with IBD will be difficult, if not impossible, to obtain because of the rarity of this complication. Finally, HHV8-induced Kaposi's sarcomas, which commonly occur after transplantation, are quite rare in patients with IBD (only one case in a CESAME patient exposed to thiopurines, corresponding to an incidence rate of 1/16 000 patient-years, unpublished data).
Other cancers
Thiopurines specifically promote acute myeloid leukaemia and severe myelodysplastic syndrome, and this is attributed to a defective DNA mismatch repair in blood cells that escape the cytotoxic effect of drugs.45 Although rare, this complication occurs significantly in patients with IBD exposed to thiopurines (CESAME cohort, unpublished data).
Although the overall excess of cancers in patients with IBD exposed to immunosuppressants seems negligible46 or low,9 many types of solid cancers have been shown to be promoted after transplantation.7 This is notably the case for melanomas and liver, urinary tract and thyroid cancers; by contrast, the incidence of de novo breast cancer is unchanged, and that of colorectal cancer (CRC) slightly or not increased after transplantation.7 Only large cohorts stratified for immunosuppressant use can answer the question about cancers in the field of IBD. In the CESAME cohort, a trend towards an excess of urinary tract cancers in patients exposed to thiopurines was seen (unpublished data).
Special contexts
Through their potential to accelerate tumour progression, immunosuppressive drugs can rapidly transform a latent cancer into a life-threatening cancer. Screening for clinically silent solid organ malignancies is now carried out before transplantation. As an illustration, silent cancers were detected in 15% of candidates for heart transplantation using extensive investigations that included a CT scan of the thorax and abdomen and screening for breast cancer in women and prostate cancer in men.47 This prevalence was significantly higher in patients older than 55 years and therefore screening could be considered for use in older patients with IBD with a late-onset disease before starting immunosuppressive therapy.
When a cancer is diagnosed in patients with IBD who are receiving immunosuppressants, practitioners should agree to a break in the immunosuppressive therapy, until the cancer is controlled. Corticosteroids (themselves often part of cancer chemotherapy protocols) can be used for treating flares during this period, with anti-TNF to be considered as the second-line drug for life-threatening uncontrolled IBD. Patients with advanced metastatic cancers who are receiving palliative or ‘semi-palliative’ treatment may have by chance, at the same time, a disabling uncontrolled IBD. The choice of starting or resuming immunosuppressive therapy is here a matter of quality of life and patient priority.
What to do with immunosuppressants in patients with IBD with a personal history of controlled cancer is more problematic. After transplantation, with deep immunosuppression during the first postoperative year, the overall recurrence rate of cancer is high (>20%), and varies according to the organ affected by the tumour (table 2).48 In patients with IBD, the overall rate of recurrence with immunosuppressants seems much lower (CESAME cohort, unpublished data). Given the current lack of evidence, individual decisions must be discussed with an oncologist, taking into account the following factors: risk of recurrence of the tumour in the general population according to the histological stage and type; importance of the role of immunosurveillance8 in tumour progression, which may vary for the same organ according to histological subtypes; risk of recurrence established in other contexts of immunosuppression (table 2); need for immunosuppressive therapy for the treatment of IBD and expected effect on IBD progression.
Inflammation-related cancers
It is now established that chronic inflammation is a common and important factor in the pathogenesis of neoplasias.49 This feature also applies to intestinal inflammation.50
Colorectal cancers
Patients with IBD with extensive longstanding colitis are at higher risk of CRC than individuals in the general population.51 Chronic inflammation of colorectal mucosa, even at microscopic levels only (neutrophil infiltrate) in patients with clinically quiescent endoscopic and clinical disease,52 appears to be major driver of these inflammation-related CRCs.50 The molecular pathways underlying IBD-related cancers are not essentially different from those of sporadic cancers (mutations, loss of heterozygosity, microsatellite instability, hypermethylation),51 53 but the sequence of events leading from normal mucosa to cancer via dysplasia is radically different, if not reversed, for several mutations. A crucial point is that several cell abnormalities, including very early events of inflammation-related carcinogenesis, may affect multiple fields of cells that look fine under the microscope.54 In other words, the process of carcinogenesis may start very early in the disease, and anti-neoplastic drugs, if not given from the time of diagnosis, may suspend the process but have no guarantee of reversing it, even if these drugs achieve sustained mucosal healing.
Chemoprevention of colorectal cancer with 5-aminosalicylates (5-ASA)
Chemopreventive agents can inhibit, delay or reverse carcinogenesis, leading to a reduced incidence of cancer. 5-ASA drugs have been used as chemopreventive agents since the 1990s, when a reduction of the incidence of CRC in patients with UC exposed to 5-ASA was reported in cohort and case–control studies. An important step in the field was the publication in 2005 of a meta-analysis suggesting that a 50% reduction in the risk of CRC occurred in patients with UC exposed to 5-ASA,55 opening the door in clinical practice to the extensive use of 5-ASA as the sole agent of chemoprevention.56
Studies with negative results have been published since then, reopening the debate. An exhaustive appraisal of the pros and cons cannot be made here, and my purpose is merely to emphasise some methodological and conceptual elements. Carrying out a randomised controlled trial specifically designed to confirm this effect is not feasible as a prohibitive number of patients at risk would need to be enrolled, and so evidence can only be obtained through observational studies. Some methodological keys are necessary for proper interpretation of each study. The ideal study (table 3) should be able to adjust the impact of a drug for all factors that influence the risk of CRC in IBD but such a study will never exist. For example, one cannot imagine a sequential assessment of microscopic inflammation in huge cohorts, and it has not yet been shown that sequential assessment of clinical IBD activity is a good surrogate marker for the intensity of mucosal inflammation. By contrast, it is essential to be able to match patients for the duration and the extent of colitis (two of the four major determinants of risk with age and gender, in addition to the particular context of associated primary sclerosing cholangitis). It is also increasingly important to control for confounding by indication with a propensity score or with other more sophisticated methods. In observational studies, drugs are prescribed in light of prognostic information. As 5-ASA is intended to prevent CRC, patients with the higher risk of developing CRC probably have been treated more often in the past few years (ie, they have a higher propensity to treatment exposure) than patients with a low risk. Consequently, if no adjustment for propensity is made, an analysis comparing treated and untreated patients may lead to an apparently positive association between 5-ASA use and CRC.
To illustrate the importance of adjustments, in the case of HIV infection, treatment with combined antiretroviral agents seems to be associated with increased mortality if no adjustment is made.57 Another illustrative example of confounding by indication bias is the 10-fold increase in the risk of gastric bleeding or perforation in users of non-steroidal anti-inflammatory drugs together with gastroprotective drugs compared with users of non-steroidal anti-inflammatory drugs alone.58 Thus we should be prudent when interpreting studies with no adjustment for the disease extent or propensity to receive 5-ASA.59 In a case–control study nested in the CESAME cohort with an adjustment for propensity, we confirmed in the early 2000s a 50% significant reduction in the incidence of CRC in patients with UC exposed to 5-ASA.60 These data should be considered together with other conceptual reasons for choosing chemoprevention with 5-ASA from the diagnosis in patients with UC not restricted to proctitis (box 1). No proof of concept of efficient chemoprevention of CRC with 5-ASA in Crohn's colitis has been found and it will be difficult to obtain as extensive colitis is not common, and 5-ASA is predominantly prescribed as chemoprevention, which maximises the confounding by indication bias.
Ten reasons for considering chemoprevention of colorectal cancer with 5-acetylsalicylic acid (5-ASA) from diagnosis in patients with ulcerative colitis (UC) not limited to the rectum
There is a convincing signal of efficacy in studies with adequate methodology (see table 3).
5-ASA, in parallel, has a proper anti-inflammatory efficacy as maintenance treatment.61
There is no demonstrated extinction of the anti-inflammatory effect of 5-ASA with time.62
There is no fear of drug-promoted cancers with 5-ASA compared with all immunosuppressive drugs.
There is a potential addition of chemopreventive effects with other immunosuppressants, with a possible pharmacodynamic synergy with thiopurines.63
There is no major long-term toxicity signal with 5-ASA in those patients not initially allergic, provided that regular monitoring of renal function is carried out.
5-ASA has several molecular antineoplastic effects that might also be of interest in prevention of sporadic cancers that can arise in patients with inflammatory bowel disease.64
Colorectal cancers may occur early in the course of UC.65
Multiple areas of molecular damage of epithelial tissue have been shown in patients with UC, before the stage of dysplasia detectable by conventional microscopy.54
Colonoscopic detection of neoplastic lesions, the alternative method for reducing the incidence of advanced colorectal cancers, is poorly implemented in practicee.66
Chemoprevention of colorectal cancer and immunosuppressants
Thiopurines are mutagenic and cytotoxic drugs that are used at high doses for treating some malignancies.67 However, there is no evidence for the molecular mechanisms of thiopurine action at lower doses that might interact with inflammation-related CRC carcinogenesis,53 as is the case for 5-ASA.64 Therefore, it can be hypothesised that the potential impact of thiopurines on the risk of inflammation-related CRC in IBD relies on a non-specific anti-inflammatory effect. This working hypothesis also suggests that good candidates for the chemopreventive effect of thiopurines are those with reduced drug-induced intestinal inflammation—that is, ideally, patients with a full clinical response combined with no persistent macroscopic or microscopic inflammation of colonic mucosa.
Until the late 1990s, the use of thiopurines was restricted to the more severe cases of IBD inflammation and used in a step-up approach; this probably resulted in a ‘mean severity of inflammation’ that was more pronounced in the patients exposed to thiopurines, and in the impossibility of demonstrating the chemopreventive effect of thiopurines.9 52 68–72 This is well illustrated by the study of Matula et al who conclude that thiopurines have a neutral effect on the risk of CRC in IBD73: in their cohort, patients exposed to thiopurines were more prone to receive concurrent corticosteroids and to undergo colectomy for reasons other than neoplasia than those in the non-exposed group. The situation has changed considerably, and up to one-third of patients with CD now receive thiopurines, with no evidence for a mean worse annual clinical activity of the disease in the exposed patients.9
In this new situation, van Schaik et al reported, in a recent issue of Gut, a 10-fold reduction in the risk of CRC in patients with IBD exposed to thiopurines, and this effect was much greater than that attributed to exposure to 5-ASA.74 This is a new and important result, which may appear too good to be true, and we must remain cautious about the level of risks estimated in this paper. The authors used a private insurance medico-administrative database and some bias of entry into the system may exist. For instance, more than one-third of the cases of CRC in patients receiving no 5-ASA or thiopurines were diagnosed during the first months of follow-up, suggesting that some patients might have entered the insurance system because of symptoms related to CRC.
In the CESAME study, a threefold decrease in incident cases of advanced neoplasia in patients at high risk of CRC (longstanding extensive colitis), when exposed to thiopurines, was reported.75 The effect was not significant for patients who were not at high risk (unpublished data), which is consistent with the hypothesis that the chemopreventive effect of thiopurines is maximal in patients at higher risk of inflammation-related CRC. We await further confirmation of this potentially beneficial effect of thiopurines on the risk of CRC in patients with IBD with colonic disease, and look forward to discovering what the effect would be with anti-TNF.
Small bowel adenocarcinoma
There is accumulating evidence that longstanding (>15 year) inflammation of ileal mucosa may also lead to dysplasia and cancer via inflammation-driven carcinogenesis.76 77 The incidence of small bowel adenocarcinoma in low in the general population, but this background risk is increased 35-fold in patients with ileal CD,78 reaching the level of age-matched risk for sporadic CRC. This is a clinical problem in IBD as no method of detection exists. Removal of the organ would suppress the risk, but preventive ileal resections would be associated with an unacceptable ‘number needed to resect’. Thus chemoprevention is important, with background concepts identical to those of CRC chemoprevention. At present, only the protective effect of 5-ASA has been suggested in a case–control study.79
Anal squamous cell carcinoma
There are numerous case reports of anal SCC occurring in patients with longstanding anal lesions.80 The existence of this rare inflammation-related cancer in IBD is plausible but epidemiological data will be difficult to obtain given the rarity of the event and the absence of reference incidence rates in the general population that discriminate between perianal and anal carcinomas. The respective roles of chronic inflammation, HPV infection81 and drug-induced immunosuppression are unknown.
Cholangiocarcinoma
Individuals with primary sclerosing cholangitis comprise <5–10% of the total IBD population, but they have the highest risk level of inflammation-related CRC from the time of diagnosis together with a lifelong risk of liver and biliary tract cancer. For CRC, there is agreement that endoscopic detection and chemoprevention with 5-ASA from the time of diagnosis should be used. In addition, bile acids may be chemoprotective.82 Preventing or detecting cholangiocarcinoma is a matter for expert hepatologists, but at the moment, there is no validated detection tool or chemoprevention.83
Towards an individually tailored reduction in the risk of IBD-related cancers
Against the general background of no excess mortality from cancer in IBD (which should not be forgotten), the challenge of clinicians is to minimise the risk of inflammation-related and immunosuppressive therapy-related cancers. This may seem an impossible challenge because the drugs which can sustain mucosal healing are also responsible for drug-related cancers if administered for long periods, justifying the initiation of benefit–risk balance simulations. We may soon arrive at individually tailored strategies based on the use of intestinal inflammation surrogate markers (imaging signals and faecal markers). For instance, if a good correlation between the normal faecal levels of proteins excreted by intestinal neutrophils and the absence of significant mucosal inflammation is confirmed, then we will be able to attempt a relatively early withdrawal of successful immunosuppressants, even if it means the prompt resumption of treatment if inflammation recurs. These individually tailored approaches, based on ‘heal and run’ periods of immunosuppression followed by ‘smart pauses,’ might lead to combined minimal cumulative periods of intestinal inflammation and drug-induced immunosuppression (figure 1).
References
Footnotes
Correction notice This paper has been modified since it was first published online. All references from 57 onwards have been corrected.
Competing interests LB declares paid participation to advisory boards or educational activities for Abbott, Ferring Pharmaceuticals and MSD.
Provenance and peer review Commissioned; externally peer reviewed.