CD22 Expression Level As a Predictor of Survival in Patients (Pts) with Relapsed/Refractory (R-R) Acute Lymphoblastic Leukemia (ALL) Treated with Inotuzumab Ozogamicin (INO) in Combination with Low-Intensity Chemotherapy (mini-hyper-CVD) with or without Blinatumomab: Results from a Phase 2 Study

Background:

The combination of reduced-intensity chemotherapy with INO with or without blinatumomab improved survival of pts with R-R ALL compared to INO monotherapy. However, outcomes and survivals vary between pts and hence it is essential to investigate the predictors of response and survival in order to improve upon treatment selection in pts with R-R ALL. In the INO-VATE trial, patients with CD22 positivity of ≥90% had better outcomes in terms of duration of response and survival than those with <90% CD22 positivity when treated with INO. The aim of the analysis is to evaluate the impact of CD22 expression in predicting outcomes in pts with R-R ALL treated with mini-hyper-CVD + INO with or without blinatumomab.

Methods:

The current analysis is based on a phase 2 study at our institution investigating the combination of low-intensity chemotherapy (mini-hyper-CVD) in combination with INO with or without blinatumomab in pts with R-R ALL. Briefly, pts with CD22-positive R-R ALL were treated with low-intensity chemotherapy compared to conventional hyper-CVAD and referred to as mini-hyper-CVD (cyclophosphamide and dexamethasone at 50% dose reduction, no anthracycline, methotrexate at 75% dose reduction, cytarabine at 0.5 g/m² x 4 doses). INO was given on Day 3 of each of the first 4 cycles at 1.8-1.3 mg/m² for cycle 1 followed by 1.3-1.0 mg/m² for subsequent cycles. The protocol was amended starting from pt #68 onwards to allow for the reduction of INO dose and its fractionation into biweekly doses (0.6 mg/m² and 0.3 mg/m² during cycle 1 and 0.3 mg/m² and 0.3 mg/m² during subsequent cycles), as well as the addition of blinatumomab for up to 4 cycles after INO therapy. Pts were categorized based on CD22 expression into 2 categories: those with 70% or greater expression of CD22 and those with less than 70% expression. The cutoff of CD22 at 70% was selected by the optimal cutoff finder (Budczies et al. PloS One 2012). The outcomes and survival of these pts were compared.

Results:

Ninety-six pts were enrolled, of whom 92 pts have available CD22 expression level for analysis. The median CD22 expression was 95% (range, 14% to 100%). Seventy-seven pts had a CD22 expression ≥ 70% (Table 1). The overall response rate (ORR) was 83% in pts with high CD22 expression compared to 67% in pts with low CD22 expression (p=0.142) with 61% complete remission (CR) in pts with high CD22 expression compared to 33% in pts with low expression (p=0.048). MRD negativity at the time of CR was achieved in 6 of 9 evaluable pts with low CD22 expression (67%) compared to 31 of 61 evaluable pts with high CD22 expression (51%) (p =0.374). Overall MRD was similar in both groups (80% vs 81% in low vs high CD22 expression) (Table 2). 4/15 (27%) pts in the low CD22 expression group proceeded to ASCT compared to 42/77 (55%) pts with high CD22 expression (p=0.048). At a median follow-up of 36 months (range, 4 to 88 months), the median OS is 17 months for pts with CD22 expression ³70% and 6 months for those with CD22 expression < 70% (p=0.0004). The estimated 3-year OS rates were 37% (95% CI, 26%-49%) and 7% (95% CI, 0.5%-26%), respectively. By multivariate analysis, baseline CD22 expression below 70% was independently associated with worse survival [HR= 4.669 (95% CI= 2.187–9.967); p<0.001] (Figure 1).

Conclusions:

Our analysis identified baseline low CD22 expression level of less than 70% as an independent predictive of poor outcome in pts treated with the combination of mini-hyper-CVD with INO with or without blinatumomab. Pts with poor risk features including low CD22 expression should be considered for alternative strategies such as CAR T-cells therapies directed at other tumor targets and clinical trials