Increased risk of lymphoma among inflammatory bowel disease patients treated with azathioprine and 6-mercapto-purine.Kandiel A, Fraser AB, Korelitz BI, Brensinger C, Lewis JD. Gut 2005;54:1121-25.
Summary: The authors of this paper performed a meta-analysis to assess the frequency with which lymphoma occurs in IBD patients treated with azathioprine and 6-mercaptopurine (6-MP). (1) Only published studies which had been designed to assess this association were reviewed. Six studies from 1985 to 2001 (two of which published by the present authors) met the strict inclusion criteria. The authors used the standardized incidence ratios SIR's from each study, developing 2 SIR's retrospectively from studies in which SIR had not been calculated). This figure is based on the number of lymphoma cases observed and the incidence expected in the general population from which the cases came with adjustments for sex and age. They then explored the reliability of their pooled analysis by evaluating the heterogeneity of the studies and checked to see what happened to their pooled estimates when they eliminated one study at a time from the analysis. A total of 3891 patients with Crohn and ulcerative colitis were included (range 238-1465) with treatment ranging from just over 1 year to 4.4 years. Follow up ranged from 2.9 to 9 years. Dosage range for azathioprine was about 2.0 mg/Kg/day in the studies from which such data could be obtained. Dose of 6-MP was 12.5-100 g/day in the single study of 6-MP.
Eleven cases of lymphoma were reported in the six studies where only 2.63 were expected. The duration of therapy ranged from 4 to 94 months. There was insufficient data to calculate the total drug exposure or the doses of individual patients. The pooled data indicated that the relative risk of developing lymphoma was 4.18 with a 95% confidence interval of 2.07-7.51. The authors agreed that the size of the study would have an impact on the SIR, since one case in a small study would carry much more weight. Indeed the estimated relative risks for the individual studies ranged from 0-37.5, the latter figure occurring in the smallest study of 238 patients with 2 lymphomas. There were 9 non-Hodgkin lymphomas with an expected incidence of only 2. The studies were too heterogeneous in size to make much of this increased incidence of non-Hodgkin lymphoma. The bowel was the most common site with Hodgkin disease in two one CNS lymphoma and 3 not identified. The authors calculated the number of person years of follow up required to uncover one additional case of lymphoma. In the 20-29 year age group, with an SIR of 4, one would have to treat 4357 cases of IBD to observe one additional lymphoma per year. In older patients (70-79 years), where the incidence of lymphoma is higher, one would have to treat only 355 patients to see one additional case.
In three studies there was enough data to compare IBD patients treated with azathioprine or 6-MP to IBD patients who had not received these drugs. The SIR was only 2.92 suggesting that IBD itself carries some relative risk for lymphoma as well. The authors found no relationship between duration of therapy and lymphoma.
Comment: This well done meta-analysis raises worrisome questions for the pediatric gastroenterologist. Although the increased risk of lymphoma with these two drugs is well known in organ transplant, rheumatoid arthritis and even long ago in chronic autoimmune liver disease, the prevailing opinion is that the risk is not increased in IBD patients (2,3). There are some concerns about this study. The authors chose to ignore many studies for reasons that might not be entirely without bias. They included two of their own studies, which might skew the results somewhat. They re-evaluated data retrospectively, a practice which also carries bias. They may have excluded reasonable studies not published in English. In pediatrics, our concern always includes the impact of duration of disease, duration of therapy and total drug exposure. These data were unavailable. The authors recognize that azathioprine and 6-MP are second line drugs used in more difficult cases and correctly suggest that the patient population might have more severe disease as a cause of lymphoma rather than a drug effect. Their data from 3 studies comparing untreated IBD patients to treated IBD patients did show some increase in relative risk among the treated patients, but this data still doesn't get to the impact of disease severity on lymphoma risk. Family histories were also not available.
Large numbers of Crohn disease patients are now receiving azathioprine as adjunct therapy to infliximab infusions. I wonder whether this group of patients with severe disease should be selected for long term monitoring. One also wonders whether monitoring for imuran metabolites, routine in most centers before starting azathioprine, might shed some light on the reasons for the increased incidence of lymphoma. This data will have to be obtained in future studies.
Lymphoma is a rare tumor. Increasing its risk by a factor of 4 still leaves it a rare complication. We badly need long term follow up in the pediatric population in order to put these data into perspective for our own patients.
Judith M. Sondheimer, MD
Chief Pediatric Gastroenterology Hepatology and Nutrition University of Colorado Health Sciences Center The Children's Hospital Denver, CO
1. Kandiel A, Fraser AG, Korelitz BI, Brensinger C, Lewis JD. Increased risk of lymphoma among inflammatory bowel disease patients treated with azathioprine and 6-mercapto-purine. Gut
2. Srinivasan R, Lewis JD. Malignancies in inflammatory bowel disease. Curr Opin Gastroenterol
3. Losco A, Gianelli U, Cassani B, et al. Epstein Barr virus associated lymphoma in Crohn's disease. Inflamm Bowel Dis