Palbociclib in combination with dexamethasone, bortezomib, and doxorubicin for pediatric relapsed acute lymphoblastic leukemia Authors: Sushmitha Grama Srinivasan, Jennifer L Kamens, Lianna J Marks, Jay M S Balagtas, Jeffrey E Rubnitz, Tanja A Gruber Published: 2026 TO THE EDITOR: Children with relapsed acute lymphoblastic leukemia (ALL) have poor outcomes in the pre–chimeric antigen receptor T-cell (CAR-T) era, with a 5-year overall survival of <50% for patients who receive multiagent chemotherapy with or without transplant.1 Although the emergence of CAR-T therapy has revolutionized the field with activity in chemotherapy-refractory disease, ∼50% of the patients receiving CAR-T therapy will experience relapse.2 Given CAR-T outcomes are more favorable in the minimal residual disease (MRD) setting and the use of chemotherapy while CAR-Ts are being manufactured, there remains a need to identify effective salvage regimens for patients with relapsed ALL.2 Palbociclib is a reversible inhibitor of cyclin-dependent kinase 4 (CDK4) and the closely related CDK6.3,4 Inhibition of CDK4/6 by palbociclib arrests the proliferation of tumor cells that have a functional retinoblastoma protein by blocking its phosphorylation thus leading to cell cycle arrest in the G1 phase.3,4 Retinoblastoma protein loss is not a recurrent alteration in pediatric ALL, whereas deletions of the CDKN2A/B tumor suppressor locus frequently occur, which combined predicts a dependency on the cyclin D–CDK6 complex.5 Building on previous preclinical data supporting (1) the activity of palbociclib in ALL, (2) the activity of palbociclib in combination with bortezomib, and (3) the activity of bortezomib in relapsed ALL, we designed a phase 1 study for relapsed/refractory pediatric ALL that evaluated escalating doses of palbociclib in combination with dexamethasone, bortezomib, and doxorubicin (ClinicalTrials.gov identifiers: NCT03515200 and NCT04996160).6, 7, 8, 9, 10, 11 The protocol and all amendments were approved by institutional review boards, and written informed consent was obtained from all patients’ legal guardians. Eligibility criteria are included in the supplemental Methods. Patients received oral palbociclib alongside dexamethasone (8 mg/m2 per day by mouth divided twice daily on days 1-5 and 11-15) to maximize synchronization of G1 arrest, followed by a 48-hour washout period.6,12 Subsequently, bortezomib (1.3 mg/m2 IV on days 7, 10, 17, and 20) and doxorubicin (25 mg/m2 IV on days 7 and 17), both cytotoxic antileukemic drugs that preferentially kill cells in the S phase, were administered (supplemental Figure 1).6,13 A classical 3+3 dose escalation of palbociclib was used to determine the maximum tolerated combination dose, beginning with a starting dose that was 66% of the previously defined single-agent maximum tolerated dose of 75 mg/m2 per day for pediatric patients with progressive brain tumors.14 Four palbociclib dose levels (DLs) were tested: 50 mg/m2 per day (DL1); 75 mg/m2 per day (DL2); 100 mg/m2 per day (DL3) by mouth on days 1 to 5 and 11 to 15; and 100 mg/m2 per day by mouth on days 1 to 5, 11 to 15, and 21 to 30 (DL4). Triple intrathecal chemotherapy with methotrexate, hydrocortisone, and cytarabine was administered on day 1; patients with central nervous system disease (CNS2 or CNS3) received twice-weekly intrathecal therapy until cerebrospinal fluid was cleared of the blasts. Toxicities were graded according to the Cancer Therapy Evaluation Program (CTEP) Common Terminology Criteria for Adverse Events version 4.03. Dose-limiting toxicity (DLT) was based on nonhematologic toxicities that occurred during the chemotherapy course, with exceptions for known toxicities associated with the treatment of patients with relapsed ALL and/or toxicities that resolved to grade ≤2 within specified timelines (supplemental Methods).15 Toxicities were monitored and recorded for up to 30 days after the participant completed treatment or until they received other antileukemia therapy, whichever came first. All 3 participants in a cohort were required to have completed the DLT monitoring phase before accrual resumed. Bone marrow aspirates were done at baseline and at the end of 1 cycle. Response assessments were determined by flow cytometry (supplemental Methods). The primary objective was to define a tolerable combination of palbociclib plus chemotherapy; the secondary objective was to estimate the overall response rate after 1 cycle. From 28 March 2018 to 18 August 2023, 14 patients with relapsed B-lineage ALL (N = 13) or T-lineage ALL (N = 1) were enrolled in the study cohort (Table 1). Age at the time of enrollment ranged from 7 months to 16 years (median, 7 years [interquartile range, 3-11]); 57% of the patients (N = 8/14) were male and 43% (N = 6/14) female. In total, 79% of the patients (N = 11/14) were white. KMT2A rearrangements were the predominant molecular genotype occurring in 43% (N = 6/14) of the cohort (Table 1). All 14 patients completed 1 cycle of treatment. No DLTs were observed at any level, and there were no grade 5 events (Tables 1 and 2). Frequent adverse grade 3 to 4 nonhematologic events across all 3 DLs include fever and neutropenia (50% [N = 7/14]), catheter-related infection (36% [N = 5/14]), aspartate aminotransferase elevation (21% [N = 3/14]), and hypokalemia (29% [N = 4/14]; Table 2). The overall response rate was 21% (N = 3/14) with 1 partial response at DL2, 1 complete morphologic remission with MRD by flow cytometry at DL3, and 1 MRD-negative complete remission at DL4 (Table 1). Table 1.Study cohortDLPatient IDAge∗PhenotypeDisease status†MolecularDLTBefore-study,‡ %After cycle 1,‡ %Response1100013.1B lineageEarly first relapseKMT2A::MLLT10No9084NR100021.8B lineageEarly first relapseKMT2A::AFF1No9098NR1000316.2B lineagePrimary refractoryP2RY8::CRLF2No9474NR2100047.7B lineageSecond relapseIL7R, PAX5No92100NR100054.3B lineageEarly first relapseETV6::RUNX1, TP53No6439NR100067.4B lineageEarly first relapseCREBBP, KRASNo966.7PR3100073.1B lineageLate first relapseKMT2A::AFF1No883.7CRM100086.2B lineageThird relapseNot availableNo4741NR100097.5B lineageEarly first relapseBCR::JAK2, IKZF1delNo9563NR4100100.6B lineageEarly first relapseKMT2A::MLLT10No4625NR100117.4B lineageThird relapseKMT2A::MLLT1No100<0.01CR1001213.9B lineageEarly first relapseKMT2A::AFDNNo7270NR1001312.4T lineageSecond relapseTP53R248Q, NOTCH1G2328fs∗25No2669PD1001411.5B lineagePrimary refractoryTCF3::HLFNo1318NRCR, complete response MRD negative; CRM, complete response morphologic (<5% blasts by flow); NR, no response; PD, progressive disease (deterioration of initial disease status); PR, partial response (>50% reduction and <25% blasts).∗Age in years at the time of enrollment.†Early first relapse (relapse <6 months from the end of therapy) and late first relapse (relapse ≥6 months from the end of therapy).‡Bone marrow involvement as determined by flow cytometry before and after a single cycle of therapy on the study. Study cohort CR, complete response MRD negative; CRM, complete response morphologic (<5% blasts by flow); NR, no response; PD, progressive disease (deterioration of initial disease status); PR, partial response (>50% reduction and <25% blasts). Age in years at the time of enrollment. Early first relapse (relapse <6 months from the end of therapy) and late first relapse (relapse ≥6 months from the end of therapy). Bone marrow involvement as determined by flow cytometry before and after a single cycle of therapy on the study. Table 2.Nonhematologic grade 3 to 4 adverse eventsAdverse eventDL1 (N = 3), n (%)DL2 (N = 3), n (%)DL3 (N = 3), n (%)DL4 (N = 4), n (%)Combined (N = 14), n (%)Grade 3Grade 4Grade 3Grade 4Grade 3Grade 4Grade 3Grade 4Grade 3Grade 4Blood and lymphatic system (fever and neutropenia)3 (100)0 (0)2 (67)0 (0)1 (33)0 (0)1 (33)0 (0)7 (50)0 (0)Gastrointestinal (diarrhea)1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0)General disorders and administration (infusion-related reaction)1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0)Infections and infestations Catheter-related infection3 (100)0 (0)1 (33)0 (0)0 (0)0 (0)1 (33)0 (0)5 (36)0 (0) Lung infection1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0) Upper respiratory infection0 (0)0 (0)1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0)Investigations ALT1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)1 (33)0 (0)2 (14)0 (0) AST1 (33)0 (0)1 (33)0 (0)0 (0)0 (0)1 (33)0 (0)3 (21)0 (0)Metabolism and nutrition Hyperuricemia0 (0)0 (0)1 (33)0 (0)0 (0)0 (0)1 (33)0 (0)2 (14)0 (0) Hypokalemia3 (100)0 (0)0 (0)1 (33)0 (0)0 (0)1 (33)1 (33)4 (29)2 (14) Hyponatremia0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)1 (33)0 (0)1 (7)0 (0) Hypophosphatemia2 (67)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)2 (14)0 (0)Musculoskeletal connective tissue (pain in extremity)1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0)Respiratory disorders Dyspnea2 (67)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)2 (14)0 (0) Hypoxia0 (0)0 (0)1 (33)0 (0)0 (0)0 (0)0 (0)0 (0)1 (7)0 (0)Vascular disorder (hypertension)2 (17)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)2 (14)0 (0)The highest grade for adverse events is shown.ALT, alanine aminotransferase; AST, aspartate aminotransferase. Nonhematologic grade 3 to 4 adverse events The highest grade for adverse events is shown. ALT, alanine aminotransferase; AST, aspartate aminotransferase. Overall, the regimen was well tolerated with no unexpected adverse events. The most frequent events included fever and neutropenia and catheter-related infections, which are common complications in the relapsed pediatric ALL setting. Both of the complete responses occurred in patients carrying KMT2A rearrangements consistent with preclinical studies supporting a requirement for CDK6 in KMT2A-rearranged leukemia.16 The response rate observed in this study is lower than that reported by the Children’s Oncology Group for AINV18P1, which evaluated 21 days of palbociclib at 50 mg/m2 per day in combination with a standard 4-drug induction in 12 patients with relapsed ALL.17 Our study had a greater cumulative exposure to palbociclib over the course of the treatment cycle (2000 mg/m2 vs 1050 mg/m2) but a less intensive chemotherapy backbone with lower amounts of anthracyclines (50 mg/m2 vs 240 mg/m2) and steroids (400 mg/m2 prednisone equivalent vs 1120 mg/m2) and no asparaginase.17 The molecular and/or cytogenetic phenotype of patients enrolled on AINV18P1 was not reported, precluding an assessment of potential molecular biomarkers of response to determine whether a clinical signal for KMT2A-rearranged disease was seen similar to our study. In summary, this study demonstrates that palbociclib at 100 mg/m2 per day can be combined safely with multiagent chemotherapy in children with relapsed/refractory ALL. Despite the low-intensity chemotherapy backbone used in this trial, we found evidence of antileukemic activity in the KMT2A-rearranged setting for patients receiving palbociclib at 100 mg/m2, including an MRD-negative complete remission after a single cycle in a patient enrolled at the third relapse (Table 1). Future studies should prioritize biomarker-enriched enrollment, particularly in KMT2A-rearranged disease, and incorporate correlative analyses of cell-cycle dynamics to identify determinants of response. Conflict of interest disclosure: The authors declare no competing financial interests.