Advertisement

Predictors of relapse and efficacy of rituximab in immune thrombotic thrombocytopenic purpura

Lova Sun, Johnathan Mack, Ang Li, Justine Ryu, Vivek A. Upadhyay, Lynne Uhl, Richard M. Kaufman, Christopher P. Stowell, Walter S. Dzik, Robert S. Makar and Pavan K. Bendapudi

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    Kaplan-Meier curve for relapse-free survival, entire cohort. Vertical red lines indicate censored patients.

  • Figure 2.

    Relapse-free survival of patients according to treatment with RTX. (A) Unadjusted Kaplan-Meier analysis for relapse-free survival at 1 year, stratified by administration of RTX. Patients begin contributing time to the RTX group starting on the day of RTX administration. The curves were compared using the log-rank test. (B) Unadjusted Kaplan-Meier analysis for relapse-free survival at 5 years, stratified by administration of RTX. Patients begin contributing time to the RTX group starting on the day of RTX administration. The curves were compared using the log-rank test. (C) Graphical representation of increasing hazard ratio of relapse rate over time for patients receiving RTX, derived from the Cox proportional hazards model.

Tables

  • Table 1.

    Demographic, clinical, and treatment data for the patient cohort

    Entire cohort (N = 124)RTX (N = 60)No RTX (N = 64)
    Study site, n (%)
     Harvard Consortium91 (73)46 (77)45 (70)
     University of Washington17 (14)9 (15)8 (13)
     Boston Medical Center16 (13)5 (8)11 (17)
    Type of presentation, n (%)
     First-time104 (84)50 (83)54 (84)
     Presenting in relapse20 (16)10 (17)10 (16)
    Sex, n (%)
     Male38 (31)18 (30)20 (31)
     Female86 (69)42 (70)44 (69)
    Age, median (IQR), y42 (31-52)41 (31-52)43 (31-53)
    ABO blood group, n (%)
     Group O64 (52)31 (52)33 (52)
     Non-group O58 (47)27 (45)31 (48)
     Unknown2 (1)2 (3)0 (0)
    Ethnicity, n (%)
     White67 (54)34 (57)33 (52)
     Nonwhite52 (42)24 (40)28 (44)
     Unknown5 (4)2 (3)3 (5)
    Laboratory values, median (IQR)
     ADAMTS13 activity, %*0 (0-0)0 (0-0)0 (0-0)
     ADAMTS13 inhibitor, BU1.4 (0.7-2.0)1.4 (0.8-2.0)1.4 (0.6-2.0)
     Hemoglobin, g/dL8.9 (7.3-10.1)8.6 (7.1-9.6)9.4 (7.4-10.8)
     Platelets, ×109/L16 (10-22)15 (10-20)17 (11-24)
     Lactate dehydrogenase, U/L1090 (742-1402)1089 (734-1504)1090 (782-1329)
     Cr, mg/dL1.0 (0.8-1.4)1.0 (0.7-1.3)1.1 (0.9-1.6)
     Total bilirubin, mg/dL2.3 (1.5-3.1)2.4 (1.5-3.1)2.3 (1.5-3.1)
     Reticulocyte, %4.6 (2.9-8.3)6.2 (3.7-9.9)3.6 (2.6-6.4)
     Day 4 platelets, ×109/L131 (68-181)111 (49-160)159 (107-216)
     Days to platelet normalization5 (4-8)6 (4-15.3)4 (3.5-6)
    Treatment
     No. (%) treated with TPE120 (97)60 (100)60 (94)
      Median (IQR) TPE procedures15 (8-23)22 (14-29)9 (6-15)
     No. (%) treated with steroids113 (91)58 (97)56 (88)
     No. (%) treated with second-line drug8 (6.5)8 (13)0 (0)
     Median (IQR) hospital stay, d12 (8-20)18 (11-27)9 (7-14)
    • * An undetectable ADAMTS13 activity level was recorded as 0%.

    • Five patients were given cyclophosphamide, 2 were given bortezomib (1 in combination with tacrolimus), and 1 patient was given vincristine.

  • Table 2.

    Outcomes

    VariableEntire cohort (N = 124)
    Number (%) who relapsed34 (27)
     Median (IQR) years to first relapse1.96 (1.05-4.07)
     Median (range) number of subsequent relapses in those who relapsed1 (1-4)
    Median (IQR) days to platelet normalization5 (4-8)
    Number (%) who died within 90 d of index presentation6 (5.5)*
    • * Calculated from follow-up data available for 109/124 patients.

  • Table 3.

    Cox multivariate model for predictors of subsequent relapse (N = 124)

    ParameterHR (95% CI)
    Presenting in iTTP relapse2.97 (1.4-6.4)
    Non-O blood group2.15 (1.06-4.39)
    Age <25 y2.94 (1.2-7.2)
    RTX*0.16 (0.04-0.70)
    • * Effect varies with time; reported HR is for day of administration. The effect of RTX was time-varying with an HR of 1.002 (95% CI, 1.0007-1.003) per day after administration.