Introduction Azathioprine (AZA) is an immunosuppressant of choice in inflammatory bowel disease (IBD). However, problems with lack of clinical response and toxicity have been encountered as the exact mode of action of AZA is not fully understood. It is thought that lymphocyte suppression may be involved. We aimed to determine if the lymphocyte count may be used to predict AZA efficacy in IBD.

Methods We conducted a retrospective study of 75 men and women with IBD (42 Ulcerative Colitis (UC), 33 Crohn's disease), maintained on AZA to investigate the relationship between the lymphocyte count and clinical course on AZA. Episodes of relapse on AZA were identified in these patients and the lymphocyte count at the visit immediately preceding the relapse recorded, as well as the relapse value itself. These were compared with values recorded during randomly selected visits with maintained remission. Other corresponding clinical data such as neutrophil, platelet, mean cell volume (MCV) counts and AZA dose were also noted.

Results Mean lymphocyte count before AZA was 2.044 (95% CI 1.861 to 2.227)×10⁹/l. This fell to 1.502 (1.445 to 1.559)×10⁹/l while patients were on AZA (p=0.000). Lymphocyte counts in patients taking AZA varied significantly according to clinical course. At visits immediately prior to or at relapse, the lymphocyte count was 1.597 (1.454 to 1.741)×10⁹/l and 1.631 (1.481 to 1.780)×10⁹/l vs 1.433 (1.367 to 1.500)×10⁹/l at visits with maintained remission (p=0.000). Of patients on AZA with a lymphocyte count below median values (1.5×10⁹/l) 57.4% were in remission compared with 37.1% of patients with higher values. (p=0.000). Similar relationships between lymphocyte counts and clinical course were noted in both UC and Crohn's disease separately. Other parameters also varied according to clinical course. MCV on treatment was significantly higher in patients in remission than prior to or at relapse (92.479 (91.662 to 93.296) fl vs 89.900 (88.145 to 91.655) fl and 89.992 (88.301 to 91.682) fl). In addition, on treatment neutrophil and platelet counts were significantly elevated prior to or at relapse. Significant differences in lymphocyte counts were also related to AZA dose. Doses >2 mg/kg were associated with lower lymphocyte counts (1.486 (1.339 to 1.633)×10⁹/l) compared with 1.695 (1.618 to 1.772)×10⁹/l on lower doses (p=0.039).

Conclusion In patients with IBD, on AZA, a lower lymphocyte count is associated with a lower incidence of disease relapse and a higher likelihood of maintained remission. Variation in other parameters such as neutrophil, platelet and MCV counts may also predict clinical course on AZA. Studies on dosing according to lymphocyte count will be a useful step towards achieving better therapeutic response on AZA.