Personalizing Therapy for Inflammatory Bowel Diseases

Ashwin N Ananthakrishnan


Expert Rev Gastroenterol Hepatol. 2013;7(6):549-558. 

In This Article

Prediction of Therapy-Related Adverse Events

Current literature also supports our ability to personalize therapy according to likelihood of adverse events. One widespread application of this practice is through TPMT testing prior to initiation of thiopurines. This allows for the identification of the 5–10% of patients who have an intermediate activity of TPMT that renders them more susceptible to drug induced leucopenia. Low activity of TPMT (genotype TPMT*3/TPMT*3) renders an individual at high risk for severe leucopenia on therapy and usually precludes use of thiopurines.[104]

The most concerning adverse events with immunosuppressive therapy are the treatment-related infections and malignancy.[13,14] Significant infections occur in 5–10% of patients treated with anti-TNF biologics.[105] One consistent risk factor that has been demonstrated is co-treatment with systemic corticosteroids. In a case–control study by Toruner et al., monotherapy with steroids, immunomodulators or anti-TNF biologics were each associated with a threefold increase in risk of opportunistic infections.[106] The risk increased substantially with combination therapy, and was greater in patients older than 50 years. Other studies have also similar identified older age to confer significantly increased risk of infections on immunosuppressive therapy.[107,108] In a tertiary referral center study by Desai et al., combination immunosuppression with a thiopurine and anti-TNF biologic was associated with a higher rate of early therapy cessation in contrast to the observed pattern in younger patients.[107] Analysis of 6273 patients enrolled in the TREAT registry additionally identified narcotic use and moderate-to-severe disease activity as risk factors for infections.[109] Larger pooled analysis of clinical trials revealed only a modest or no increase in risk of infections with infliximab alone compared with immunomodulator therapy.[15,110] A particularly severe infection is the JC-virus-mediated progressive multifocal leukoencephalopathy (PML) associated with natalizumab with an estimated risk of one in 1000.[111] The unpredictability and the severity of this rare adverse event limited widespread use of a potentially effective therapy for CD. However, all cases of PML have occurred in patients who were seropositive for the JC virus, and recent commercially available testing to assess seropositivity before treatment has allowed us to tailor natalizumab use by increasing both provider and patient confidence regarding drug safety in seronegative individuals while avoiding long-term use in patients who are seropositive. Approximately half the population tested will be seropositive for JC virus, and the risk of PML in such patients in the setting of prior immunosuppression may be as high as 8.1 out of 1000 with 2 or more years of use.[111] In patients who are seronegative, the risk even with prolonged use of natalizumab appears to be less than 0.1 out of 1000.

Non-Hodgkin's lymphoma occurs two- to five-times more commonly in patients on immunosuppressive therapy compared with the general population.[14,112,113] In a large French study by Beaugerie et al., the incidence of lymphoma with thiopurines use increased from 0 to 1.68 per 10,000 person-years in never users to 5.4 per 10,000 person-years in those over the age of 65 years.[113] A meta-analysis of six pooled studies identified a fourfold increase in risk of lymphoma in IBD patients with the estimated risk similar across thiopurines and anti-TNF biologic therapy.[112] However, a particularly aggressive variant of lymphoma that has been nearly universally fatal is the hepatosplenic T-cell lymphoma (HSTCL). While occurring with either thiopurine monotherapy or combination anti-TNF/thiopurine therapy, in a case series of 36 patients with HSTCL, over 90% were male with a median age of 23 years. The overall estimated risk of one in 45,000 could be further personalized by age and gender. The risk in men younger than age 35 years was estimated to be as high as one in 3534 with combination therapy or one in 7404 with monotherapy, substantially higher than the absolute risk in women or older men.[114]