Background/Purpose:   Physicians and patients (Pts) are hesitant to withdraw immunosuppressant (IS) in Pts in clinical remission as the consequences of this approach are unknown.

We aimed to determine the number of successful withdrawals of IS and their predictors in a large observational cohort study.

Methods:   Analysis was conducted on all Pts seen in The Lupus Clinic, from 1987-2012, in whom IS was tapered and stopped. Pts who were in clinical remission and on prednisone (P) ≤7.5mg/day were included. Tapering start was defined as the date of the visit with a decrease ≥25% in IS dose. IS Stop was the day of IS discontinuation. Study end was the date of flare or last clinic visit following IS stop.

Flare was defined as the introduction of new IS or increase of P dose for active disease. Flare was evaluated within the first 2 years from IS stop and at any time after IS stop.

Kaplan-Meier curve was used to evaluate the time to flare after IS stop. Pts who flared after IS stop were compared to Pts who did not flare (t-test and χ² test) at the time of IS tapering start and IS stop.

Covariates evaluated in the univariate analysis were: sex, ethnicity, IS, DNA antibody level and DNA antibody [yes/no], C3/C4 level and low C3/C4 [yes/no], lupus duration, age at IS taper, length of time on IS, disease activity [SLEDAI-2K, AMS year 1 before IS taper] and steroids at IS stop [yes/no]. Forced and stepwise regression models were fitted with covariates with p<0.1 in addition to age, sex and ethnicity to predict flare in Pts who discontinued IS.

Results:   Of the 1678 lupus Pts, 973 were ever on IS, 179 had tapering attempts and 99 Pts stopped IS. 91% were female and at tapering start age was 40.4±13.1 and disease duration was 11.4±9.4 years.

Of the 99 Pts, 25 flared within 2 years (16 AZA; 7 MTX and 2 MMF; p 0.31). The length of time from tapering start to IS stop was 1.8±1.8 years in the no flare and 0.9±0.9 years in the flare group; p 0.002.

46 of the 74 Pts who had not flared by 2 years had follow-up available beyond 2 years; 32 were followed beyond 3 years and 24 beyond 5 years. 17 Pts experienced a flared after year 2. Using Kaplan-Meier curve for time to flare showed that at 1, 2, 3, 4 and 5 years, the percent of Pts who flared was 17%, 30%, 46%, 49% and 51% respectively; Figure 1.

The percentage of Pts on P at the time of IS stop was greater among those who flared, 52% compared to 30%; p 0.04. At the time of IS tapering the models were not statistically significant for all studied covariates. At the time of IS stop, the results from the logistic regression showed that Pts off P are more likely not to flare; OR 2.99; 95% CI: 1.13, 7.89; p 0.03.

Conclusion:   Within 2 years, successful stopping of IS was possible in about 75% of clinically stable Pts. Half were successful within 3 year and this proportion was stable up to 5 years. At the time of IS stop, Pts who discontinued IS slowly and who were off P were less likely to flare.

Figure 1: Percent Flared since IS stop