,17 we believe that one case, in which UC developed 4.5 years after IFN-β1a treatment was discontinued, was not caused by IFN, as no evidence was provided to support this assumption of causation. The interval between the initiation of IFN therapy and development or exacerbation of UC was 7.7 ± 9.8 months (mean ± standard deviation) in 14 of the remaining 15 cases (one case lacked a detailed description).

This interval was 4.7 ± 5.2 months for the nine cases reported in Japan (Table 1) and 13.1 ± 14.1 months LDE225 clinical trial for the five cases reported in Europe and the USA (Table 2). Although these intervals appeared to be shorter in the cases reported in Japan, the difference was not significant (unpaired Student’s t-test; P = 0.06, one-tailed). If we divide the reported

cases into three groups, the interval from the initiation of IFN treatment to development of UC was 4.3 ± 4.9 months for patients with type-C chronic hepatitis (n = 11), 9.0 ± 4.2 months for those with renal cell carcinoma (n = 2), and 28.5 ± 20.4 months for those with MS (n = 4). This interval was significantly NVP-BGJ398 ic50 longer for patients with MS (unpaired Student’s t-test; P < 0.01, one-tailed). The number of patients with UC in Japan is reported to be 104 721, and the number of new cases of UC per year is approximately 5000.28 In some of these new cases, patients had long experienced the symptoms of UC, but the condition was only recently diagnosed. In 1995, Morita et al. reported the incidence of UC development as 1.95 cases per 105 person-years, and the incidence is rising in Japan.29 With 5000 new cases per year28 and the Japanese population of 200 million, the incidence of UC is about 2.5 cases per 105 person-years. Asakura et al. reported the UC prevalence in Japan to be 63.6 cases per 105 persons in 2005.30 The frequency and prevalence of UC in Japan are relatively low;29,30 thus, UC could be considered

a rare disease. Furthermore, because development or exacerbation of UC in response to IFN therapy is rare, its occurrence during or after IFN therapy may be due to random causes. Even if the data from recently published studies31,32 are sufficient to support a causal correlation between IFN treatment GBA3 and UC (or support its effectiveness for treating UC), we must compare the frequency of UC development or exacerbation between the general population (i.e. those who have not undergone IFN treatment) and patients with chronic hepatitis C, renal cell carcinoma, or multiple sclerosis who have been treated with IFN. However, it is reasonable to suspect that published cases of UC associated with IFN are in fact due to an adverse reaction to IFN for the following reasons: (i) the interval between the development of UC and initiation of IFN treatment is relatively short; (ii) in many cases, acute symptoms (e.g.