Background: Thiopurines including azathioprine (AZT) and 6- mercaptoprine (6MP) have proven efficacy in inflammatory bowel disease (IBD). However, thiopurine-related adverse drug reactions are frequent ranging from 5% up to 40%, in both a dose dependent and independent manner. The aim of our study was to assess safety of thiopurine therapy through a retrospective tunisian series.
Methods: We have conducted a retrospective study including patients treated with thiopurines from 2006 and 2012. Epidemiologic, clinical and therapeutic characteristics were abstracted from medical records. Thiopurine-related adverse effects (AE) were sought in each patient so as to determine predictive factors of thiopurine AE. Data entry and analysis were performed by spss version 21.0.
Results: We have colliged 210 patients (98 males and 112 females) of mean age of 29.8 years old. One hundred sixty-nine patients had Crohn's disease (CD), 27 had ulcerative colitis (UC) and 12 had unclassified colitis (UnC). AZT and 6MP were prescribed respectively in 98.1% and 9% of patients. During a mean follow-up period of 28.4 months, digestive intolerance of AZT was noted in 14 patients after 5 months of treatment leading to a switch to 6MP in 10 patients. Immunoallergic reactions occurred in 8 patients (acute pancreatitis (n=5), cutaneous rash (n=3)). Hematologic toxicity was seen in 25 patients after 20 months (2–80) of treatment: lymphopenia (n=19), neutropenia (n=11), anemia (n=15) and thrombopenia (n=11). Treatment withdrawal was decided in 32 patients (15.2%) because of adverse events (n=15) and lack of response (n=17). Six patients had hepatic toxicity: cholestasis at 2 to 3 ULN resulting in a dose reduction in 3 patients. Acute hepatic cytolysis at 3 to 9 ULN occurred in 4 patients after ruling out a viral origin. Regenerative nodular hyperplasia was seen in only 1 patient. There have been one case of acute myeloid leukemia diagnosed 3 months after AZT onset. In univariate analysis, corticosteroid-dependent patients had significantly less AE than other patients (30% vs 70%, p=0.008). Patients with corticosteroid-resistance profile had les AE with trend to marginal significance (6% vs 94%, p=0.08). Patients who had extensive ileal involvement and who were more than 20 years old at disease onset developed digestive intolerance less rapidly (respectiveley p=0.06 and p=0.04). Immunoallergic reactions seem to occur less commonly among patients who had been previously treated with corticosteroids (p=0.09).
Conclusions: Overall, use of thiopurines in patients with IBD is safe. Hematologic and hepatic toxicities are the most common side effects. Clinicians should cconsider these side effects so as to optimize thiopurine therapy in IBD patients.