Table 3.

Summary of multiple regression analyses for durability of response

VariableHR (95% CI] for relapse risk
Pretreatment variables onlyPretreatment variables plus response depth (PR vs CR) at 9 moPretreatment variables plus MRD status (pos vs U) at 24 mo
All patients (n = 323)400 mg monotherapy (n = 250)All patients (n = 311)400 mg monotherapy (n = 242)All patients (n = 187)400 mg monotherapy (n =)159
Max node size, cm
 ≥5 to <102.5 (1.7-3.7)2.4 (1.5-3.7)2.1 (1.4-3.3)2.1 (1.3-3.3)2.9 (1.7-5.1)2.8 (1.6-5.1)
 ≥102.8 (1.7-4.6)2.7 (1.5-4.8)2.3 (1.3-3.9)2.4 (1.4-4.3)3.6 (1.6-7.9)3.0 (1.4-6.8)
Prior therapies
 2 to 3NINININININI
 >3NINININININI
Fludarabine refractory disease1.5 (1.1-2.2)NI1.5 (1.1-2.2)NI1.0 (0.6-1.8)*NI
Prior BCRi exposureNINININININI
BCRi refractory disease2.5 (1.5-4.1)† 1.9 (1.1-3.3)2.3 (1.4-3.9)1.9 (1.1-3.2)3.0 (1.6-5.8)2.6 (1.4-5.0)
TP53 mutation and/or 17p deletion1.8 (1.2-2.5)NI1.8 (1.3-2.7)NI1.9 (1.0-3.4)NI
13q deletionNINININININI
NOTCH1 mutation1.8 (1.1-3.0)NI2.0 (1.2-3.3)NI2.6 (1.1-6.2)NI
IGHV unmutated2.1 (1.0-4.3)NININININI
Monotherapy vs combination0.8 (0.4-1.5)
PR/nPR vs CR/CRi at 9 mo2.1 (1.1-4.0)3.3 (1.0-10.5)
PB MRD-pos vs U-MRD at 24 mo3.0 (1.7-5.2)3.4 (1.9-6.3)
  • All variables included in univariate analyses were tested in the multiple regressions. Covariates selected by stepwise regression are reported, and the effects of monotherapy vs combination, PR/nPR vs CR/CRi at 9 mo, and PB MRD-pos vs U-MRD at 24 mo were added as a variable of interest after model selection, separately. Those bolded are significant; and italicized if confirmed as significant in a sensitivity analysis that only included patients in whom all data for all variables were informative (supplemental Table 6).

  • NI, not included in final model; U-MRD, minimal residual disease not detected by flow cytometry at 10−4 leukemic cell cutoff; MRD-pos, detectable MRD; —, not tested/not applicable; PB, peripheral blood.

  • * Significant pretreatment variable that is not statistically significant when a posttreatment response variable is included in the model.

  • In these 3 instances, prior BCRi exposure had higher priority in the model outputs of the sensitivity analyses than BCRi-refractoriness, and was statistically significant.

  • After pretreatment variables of significance in multiple regression analyses were identified, the treatment variable of venetoclax monotherapy or combination with rituximab was added into the Cox regression model.

  • After pretreatment variables of significance in multiple regression analyses were identified, the posttreatment response variable of either PR or CR at 9 months or MRD-pos or U-MRD status at 24 months was added into the Cox regression model.