Venetoclax in Combination With Pediatric-Inspired Chemotherapy in Adults With Newly Diagnosed Acute Lymphoblastic Leukemia: Results of a Phase I Trial

Abstract

While the outcomes of acute lymphoblastic leukemia (ALL) treatment in young children are excellent with > 90% long-term event-free survival (EFS) [1], outcomes among adults have continued to lag behind. Even with the utilization of pediatric-inspired chemotherapy (PIC) regimens, there remains a significant unmet need for 25%–40% of adult ALL patients who are not cured with upfront chemotherapy [2, 3]. Venetoclax is a B-cell lymphoma 2 (BCL-2) inhibitor approved for the treatment of acute myeloid leukemia (AML) and chronic lymphocytic leukemia (CLL) in combination with other agents. Preclinical data has long indicated that ALL blasts express high levels of BCL-2 and are sensitive to its inhibition [4]. Based on this data, several clinical trials have incorporated venetoclax in combination with different chemotherapy regimens for relapsed/refractory (R/R) ALL [5-7] as well as in upfront treatment [8] with encouraging results. These reports prompted the question of whether venetoclax can be safely added to a PIC regimen and whether that approach can improve rates of measurable residual disease-negative complete remission (MRD-CR) rates compared to historical treatments.

We conducted a phase I clinical trial of oral venetoclax in combination with asparaginase-containing PIC in adult patients with newly diagnosed ALL (NCT05386576). The study was approved by the MSKCC Institutional Review Board and conducted in accordance with the Declaration of Helsinki. The study design was a 3 + 3 dose de-escalation design with 2 dose levels. The total study accrual goal was 12 patients. Inclusion criteria included age 18–60 at the time of registration, ECOG performance status 0–2, and adequate organ function. Further details on the study design, definitions of dose limiting toxicities, and treatment plan are provided in the accompanying appendix.

The chemotherapy backbone regimen used in this trial is identical to the one utilized in the published MSKCC PIC protocol (Figure S1) [3]. Venetoclax was administered in the following fashion in the initial protocol: During Induction 1, 100 mg on day 5, 200 mg on day 6, and 400 mg on days 7–28. In patients with Grade 4 neutropenia (Absolute Neutrophil Count (ANC) < 500/μL) with fever or Grade 4 thrombocytopenia (platelet count < 25 000/μL) after day 14, venetoclax was held until ANC recovered to ≥ 1000/μL and platelets recovered to ≥ 50 000/μL. Patients proceeded from Induction I to Induction II on day 35–43 regardless of count recovery, unless the patient demonstrated marrow aplasia in two subsequent biopsies, in which case they were removed from the study (this did not occur to any patient). During Induction II, venetoclax was given at 400 mg during days 1–14 and 29–42. In patients with ANC < 500/μL with fever, ANC < 100/μL irrespective of fever, or platelet < 25 000/μL after day 29, venetoclax was held until ANC recovered to ≥ 1000/μL and platelets recovered to ≥ 50 000/μL. Due to observed prolonged cytopenias, the protocol was amended twice to change the duration of venetoclax dosing (Figure S1). The first amendment, which was applied beginning with the 4th patient, shortened the duration of venetoclax during Induction II to 400 mg on days 1–7 and 29–35. The second amendment, which was applied beginning with the 9th patient, additionally shortened the duration of venetoclax during Induction I to 14 days if the day 14 bone marrow aspirate showed MRD- CR/CRi and a hypocellular marrow. The use of venetoclax beyond induction is also shown in Figure S1.

We utilized a historical cohort of patients treated on MSK protocol 12-266 as a comparator cohort [3]. This consisted of 22 patients treated at our institution with the same PIC backbone.

A total of 12 patients were enrolled in this trial, as specified in the protocol (Table 1). Their median age at diagnosis was 39 years (IQR: 31–53). Half of the patients (n = 6, 50%) had B-cell ALL, while 6 (50%) had T-cell ALL/LBL, and two patients had early T-cell precursor (ETP) subtype. The baseline characteristics of the study cohort were statistically similar to the comparison cohort of 22 patients treated with PIC without venetoclax (Table 1). Table S1 shows the number of doses of venetoclax that patients received during Induction I and Induction II at each phase of the protocol.

TABLE 1. Patient characteristics.

Characteristic	PIC + Ven (study cohort) N = 12a	PIC (comparator cohort) N = 22a	p b
Age group at diagnosis			0.8
< 40	6 (50%)	12 (55%)
> =40	6 (50%)	10 (45%)
Sex			0.6
Female	3 (25%)	3 (14%)
Male	9 (75%)	19 (86%)
Body mass index (BMI) category at diagnosis			0.2
Normal (BMI 18.5–24.9)	1 (8.3%)	8 (36%)
Overweight (BMI 25–29.9)	6 (50%)	8 (36%)
Obese (BMI 30+)	5 (42%)	6 (27%)
Immunophenotype			0.3
B-cell	6 (50%)	17 (77%)
Philadelphia-like diseasec	2	1
T-cell	6 (50%)	5 (23%)
Early T cell precursor	2	2
CNS status at diagnosis			0.8
1	10 (83%)	19 (86%)
2	1 (8.3%)	2 (9.1%)
3	1 (8.3%)	0 (0%)
Indeterminate	0 (0%)	1 (4.5%)
Unequivocal Extramedullary disease (non CNS)	8 (67%)	8 (36%)	0.091
Cytogenetic risk category			0.072
Favorable	1 (8.3%)	0 (0%)
Intermediate	5 (42%)	5 (23%)
Unfavorable	4 (33%)	4 (18%)
Not evaluable/classified	2 (17%)	13 (59%)

• Abbreviation: PIC, pediatric inspired chemotherapy.
• ^a Median (Q1, Q3); n (%).
• ^b Wilcoxon rank sum test; Pearson's Chi-squared test; Fisher's exact test.
• ^c Of the 2 patients in the study cohort, one had a CRLF gene arrangement, while the other was suspected to have Ph-like ALL with an IKZF1 deletion, although confirmatory genetic abnormalities could not be assessed due to limited testing availability.

During the analysis period of Induction I and II, all patients in the trial experienced hematologic adverse events, as expected. Figure 1 shows the duration of Grade 4 neutropenia (ANC < 500/μL) and Grade 3+ thrombocytopenia (platelet count < 50 000/μL) in the trial patients and the historical comparator cohort. During Induction I, patients in the trial had a longer duration of Grade 4 neutropenia than the comparator cohort (median 21 vs. 10 days, p = 0.024). During Induction II, trial patients had a longer duration of Grade 4 neutropenia (median 34 vs. 9 days, p < 0.001) and Grade 3–4 thrombocytopenia (median 47 vs. 11 days, p < 0.001). Figure S2 shows the hematologic parameters at the start of Induction II. Figure S3 shows the duration of cytopenias after amendments 1 and 2; these differences were not statistically significant. The median duration of Induction I was 49 days on the trial versus 39 days on the comparator group (p = 0.10), and the median duration of Induction II was 93 days on the trial versus 74 days on the comparator cohort (p = 0.013) (Figure S4). As shown in Table S2, the duration of Induction II was numerically shorter after the protocol amendments, although that was not statistically significant.

FIGURE 1

Open in figure viewerPowerPoint

Duration of cytopenias during Induction I and II in the study cohort and comparator cohort. PIC, pediatric-inspired chemotherapy.

Grade 4 non-hematologic adverse events were as follows (Table S3): 4 patients experienced hypertriglyceridemia, 1 patient experienced alkaline phosphatase elevation attributed to asparaginase, 1 patient experienced a lung infection, and 1 patient experienced a fungal sinusitis. Notable Grade 3 events were as follows: Nine (75%) patients experienced febrile neutropenia, 4 (33%) experienced sepsis, 1 (8%) experienced pancreatitis requiring hospitalization, and 1 (8%) experienced an upper gastrointestinal hemorrhage. No patients experienced a DLT.

Table 2 shows the response assessments at the end of Induction I and II. At the end of Induction I, 7 (58%) patients had achieved an MRD- CR/CRi. This number rose to 9 (75%) at the end of Induction II; one T-ALL patient (treated before the amendments) had persistent low-level extramedullary disease only; therefore, he was classified as no response. One ETP-ALL patient (treated after amendment 2) had MRD+ CRi at the end of Induction II, and one T-ALL patient (treated after amendment 2) had indeterminate disease status due to a mildly FDG-avid lymph node on PET scan that was not biopsied at this treatment stage. These results were statistically similar to what was observed in the comparator cohort. There were no statistical differences in response among patients treated before and after the amendments (Table S4). Figure S5 shows the depth of response by bone marrow abnormal leukemic blast percentage measured with flow cytometry at three timepoints. Of note, both patients with Ph-like disease achieved MRD- CR/CRi by the end of Induction II, and both patients with ETP ALL achieved CR by the end of Induction II (one MRD− and one MRD+).

TABLE 2. Response assessments.

Characteristic	PIC + Ven N = 12a	PIC N = 22a	p b
Disease status at end of induction I			> 0.9
MRD− CR/CRi	7 (58%)	14 (64%)
MRD+ CR/CRi	4 (33%)	7 (32%)
Active disease	1 (8.3%)	1 (4.5%)
Disease status at end of induction II			0.3
MRD− CR/CRi	9 (75%)	18 (82%)
MRD+ CR/CRi	1 (8.3%)	4 (18%)
Active disease	1 (8.3%)	0 (0%)
Indeterminate	1 (8.3%)	0 (0%)

• Abbreviation: PIC, pediatric inspired chemotherapy.
• ^a n (%).
• ^b Fisher's exact test.

Figure S6 shows the overall survival (OS) and EFS curves for patients on the trial and the comparator group. At 12 months, there was 90% OS (95% CI: 73%–100%) in the trial (median follow-up of 20 months) and 100% in the comparator group (median follow-up of 83 months) (p = 0.081). At 1 year, EFS was 72.2% (95% CI: 50%–100%) in the trial versus 86.4% (95% CI: 73%–100%) in the comparator group (p = 0.36).

Our study demonstrates that venetoclax can be combined safely with pediatric-inspired chemotherapy in adult patients with untreated ALL, but the duration of venetoclax administration should be carefully assessed to minimize prolonged cytopenias and infections. Our amendments, which shortened the duration of venetoclax, led to a numerically shorter duration of neutropenia and thrombocytopenia, but this did not reach statistical significance, possibly due to inadequate power. Recently, a report of 24 patients in a phase I trial adding venetoclax to the backbone of a pediatric multiagent chemotherapy regimen (CALGB 10403) was published [9]. In that trial, which administered venetoclax for 14 days during induction and consolidation, prolonged cytopenias do not appear to be an issue; the most likely explanation for the difference in cytopenias is the shorter duration of venetoclax in that study, although the somewhat different patient age (median age 31 vs. 39 years in our study) could have also contributed.

Regarding efficacy, a high rate of responses was observed in our study, with all patients in high-risk subgroups (Ph-like and ETP) achieving CR/CRi after Induction II. We note that T-ALL can be dependent on both BCL2 and BCLXL signaling [10], possibly limiting the benefit of BCL2 inhibition. This might explain why all three patients who did not achieve MRD- CR after Induction II were T/ETP-ALL patients. Taken together with the results of the study incorporating venetoclax on a CALGB 10403 backbone, our study suggests that venetoclax can be incorporated into future protocols for subgroups at risk for chemotherapy resistance. Combinations with reduced-intensity chemotherapy such as mini-CVD might be attractive for older patients as currently being explored.