Significant changes from the Original Protocol to Amendment 1 are: • Added PK sample collection times of 2 and 4 hours after the end of carfilzomib infusion • Germline DNA collection time moved to the end of Induction • Pharmacodynamic sample collection added on Day 29 of Induction.

• Added details of MRD specimen collection, including the type of specimen to be submitted or the timing of the specimen collection • Added detailed information regarding timing of pharmacodynamics specimen collection • Corrected inconsistencies within the protocol and between the protocol and the laboratory manual regarding the proper day for collection of blood or saliva to serve as a source of normal (non-tumor) DNA from subjects who have consented to the optional genomic studies • Removed ≥ Grade 3 peripheral neuropathy or neuropathic pain from the definition of dose limiting toxicity, due to its strong association with vincristine and rare association with carfilzomib • Corrected the dose modification of vincristine in the setting of constipation, ileus, or typhlitus to reflect modification at ≥ Grade 3, rather than > Grade 3 • Added information about PRES and the required dose modification of carfilzomib when PRES is suspected • Updated the options for pregnancy testing during screening to include a urine pregnancy test, to include a urine pregnancy test, which is less invasive and more readily available at pediatric facilities • Increased the duration of required contraception to incorporate the best practices for all drugs in the treatment regimen • Modified the definition of permanent sterilization and revised the list of highly effective methods of contraception, to be consistent with recent guidance • Added language to allow for higher total bilirubin level at screening, for subjects with a diagnosis of Gilbert syndrome • Added language to clarify that the restricted use of antineoplastic agents prior to enrollment only applies to the use of agents with therapeutic intent.

• Clarify the target number of participants to be enrolled in each age group • Change requirement for medical monitor approval with requirement to contact medical monitor • Remove the requirement for bone marrow biopsy during screening Revise Section 8.4 (Dose Modification Guidelines) to o Add text clarifying that the carfilzomib dose modification guidelines also apply to the Optional Consolidation Cycle o Add dose modification guidelines for daunorubicin • Add Section 8.6 (Product Complaints) • Revise Section 9 (Supportive Care Requirements and Guidelines) to o Clarify that fungal prophylaxis must be provided during periods of neutropenia o Correct definition for fever and neutropenia o Update Section 9.14 (Contraception) • Update Section 12.6 (Pregnancy and Lactation Reporting) • Add guidelines for taking temperature and blood pressure of participants • Add Section 15.7 (Publication Policy) • Update Section 3 (Background Information) • Make minor text clarifications, additions, and edits throughout the protocol.

The protocol is amended to: • Add a second dose escalation portion to evaluate carfilzomib with a different chemotherapy backbone (VXLD) of vincristine, dexamethasone, PEG-asparaginase, and daunorubicin Rationale: Toxicity related to the R3 induction chemotherapy backbone of dexamethasone, mitoxantrone, PEG-asparaginase, and vincristine is suspected to have limited the ability to dose escalate carfilzomib. A second dose escalation using the less toxic VXLD backbone is being added to the protocol to ensure the optimal dose of carfilzomib is selected before moving on to a Phase 2 study. • Remove the Phase 2 portion of the study Rationale: The Phase 2 portion of the study will instead be conducted with a stand-alone protocol treating participants at the defined MTD. • Revise eligibility criteria to o Include enrollment of participants aged 21 years or younger at the time of initial ALL diagnosis and limiting enrollment to those aged > 1 year at the time of study treatment initiation o Include the enrollment of participants with any first relapse of T-ALL, removing the limitation to those with an early relapse o Modify definition of adequate liver function o Specify grade of pancreatitis as an exclusion criterion o Clarify evidence for bacteria or fungal infection o Clarify antineoplastic agents that are included in prior therapy restrictions.

The protocol is amended to: • Revise inclusion criteria regarding hepatic function Rationale: Provide clarity that participants with hyperbilirubinemia due to Gilbert syndrome are only eligible if they have a direct bilirubin ≤ 1.5 x institutional upper limit of normal. • Modify carfilzomib infusion time from "approximately 30 minutes" to "30 ± 5 minutes" Rationale: Provide clarity on the infusion time ensuring all participants are exposed to the same range of IV infusion time. • Add language within the Induction Cycle describing the dose level lists with the carfilzomib dose for each dose escalation and specify that during Dose Escalation 2, participants will receive 20 mg/m2 of carfilzomib on Day 1 of the Induction Cycle. Rationale: This language was erroneously omitted during Amendment 3. • Update Study 20140106 Induction Cycle Dose Escalation 2 (VXLD) road map with actual dose administered placeholders to match the VXLD backbone. Rationale: Dexamethasone, vincristine, and PEG-asparaginase administration placeholders were not correctly placed within the appropriate study day. • Administrative updates, editorial changes, and style and formatting revisions have been made to improve clarity and consistency throughout the document. Significant changes are described in the table below. Detailed, changed text is displayed for first major occurrence only. Changes in sections of the protocol body were also made in the protocol synopsis and elsewhere in the document, as applicable. Deletions of text are presented in strikethrough format. Added text is presented in bold format.

The protocol is amended to: • Modify the language for dexamethasone dose instruction in the VXLD regimen from 6 mg/m2 twice daily to 6 mg/m2 per day BID – given orally (3 mg/m2 per dose given twice daily). Rationale: Update the language describing required dexamethasone dosing instructions to ensure clarity and remove ambiguity. • Update the posterior reversible encephalopathy syndrome (PRES) language. Rationale: Update PRES language to ensure alignment between study protocol and the carfilzomib Investigator Brochure v17.1. • Update PRES management recommendation with a no “re-challenge requirement”. Rationale: Prohibit reintroduction of carfilzomib if PRES is identified in a participant at the request of the Global Safety Team (GST). • Administrative updates, editorial changes, and style and formatting revisions have been made to improve clarity and consistency throughout the document. Significant changes are described in the table below. Detailed, changed text is displayed for first major occurrence only. Changes in sections of the protocol body were also made in the protocol synopsis and elsewhere in the document, as applicable. Added text is presented in bold format.

The protocol is amended to: • Modify the stopping criteria for the Bayesian algorithm Rationale: Changes to language on how the Bayesian output will be used to determine the maximum tolerated dose in order to trigger the end of the study. • Update the cohort safety review committee language. Rationale: Update increases the discretion of the cohort safety review committee to recommend the maximum tolerated dose. • Add 2 additional dose levels in dose escalation 2 Rationale: Accumulating pharmacokinetic data in children indicates drug exposure was approximately 40% lower than that of adults at the same dose and schedule; therefore, current dose escalation plan updated to include 2 higher dose levels. • Minor modifications to the definition of dose limiting toxicities that are based on biochemical adverse events without clinical manifestations • Administrative updates, editorial changes, and style and formatting revisions have been made to improve clarity and consistency throughout the document. Significant changes are described in the table below. Detailed, changed text is displayed for first major occurrence only. Changes in sections of the protocol body were also made in the protocol synopsis and elsewhere in the document, as applicable. Added text is presented in bold format.

The protocol is amended to: • To add hepatitis B testing at screening for all participants and at the safety follow up for participants with positive hepatitis B serology at screening or past history of HBV infection. Participants with hepatitis B infection with positive hepatitis B DNA are excluded from the study. In addition, guidance was provided for the monitoring and management of HBV infections. • To allow enrollment of participants with first refractory bone marrow relapse occurring any time after original diagnosis after achieving a CR (ie, ≥ 1 failed attempt to induce a second remission) in the study. • To indicate that prior therapy with inotuzumab within 36 days (3 antibody half-lives), rather than 30 days, is exclusionary. • To indicate that next generation sequencing will be used to assess MRD status and lymphoblasts at the end of the Induction Cycle. • To indicate that the degree of proteasome inhibition will only be assayed in whole blood. • To clarify that participants should be in a rested and calm state before blood pressure measurements but are not required to be supine. • To clarify the reporting requirements for AEs that are laboratory findings that do not result in a clinical action or alter treatment. • To clarify that for the Cohort Safety Review Committee and the Data Monitoring Committee as-is snapshots will be used for the analyses. • To clarify that for the final analyses the database will be cleaned, processed and a locked database used in the analysis. • Editorial changes (including typographical, grammatical, and formatting) have been made throughout the document.

The protocol was amended as follow: • Addition of the phase 2 part to the existing phase 1 study to allow currently participating sites to seamlessly continue to enroll participants without a break in availability of this therapy for participants that may benefit from carfilzomib combined with vincristine, dexamethasone, PEG-asparaginase, and daunorubicin (VXLD). • The primary objective of the phase 1 part of the study has been amended to clarify that in addition to determining the maximum tolerated dose, the objective is also to recommend a phase 2 dose of carfilzomib with induction therapy. • Typographical and formatting changes have been made throughout the document.

• Details on non-investigational products and other protocol-required therapies were added to the phase 2 portion of the study for clarity. • The recommended dosing of trimethoprim sulfamethoxazole was added along with suggested alternatives for contraindication or significant concerns of bone marrow suppression, with a requirement to include the rationale for the use of an alternative agent in the case report form. • Clarification was added that adjustment of the carfilzomib dose is required for certain toxicities. • Dose modification guidelines, including adding guidance for progressive multifocal leukoencephalopathy and resuming carfilzomib with controlled hepatitis B reactivation, were aligned with those of the carfilzomib program for consistency across studies. • Supportive care requirements and guidelines were updated and guidelines for mucositis was added based on the recommendations of the steering committee. • The most impactful difference in participant population between phase 1 and phase 2 was clarified. • A correction was made to the inclusion criterion that adequate cardiac function is defined as shortening fraction of ≥ 30% > not 30%. • A correction was made to the inclusion criterion that life expectancy must be  6 weeks not ≥ 6 weeks. • Clarification was added to the exclusion criterion that participants should be excluded for intolerance, hypersensitivity, or inability to receive any of the chemotherapy components of the VXLD regimen (or acceptable substitutes as defined in the protocol). • Corrections were made to the exclusion criteria that participants should refrain from becoming pregnant or breastfeeding and must use contraception for 6 months after the last dose of any study treatment or for 12 months after the last dose of cyclophosphamide if used. • A recommendation was added that oral hormonal contraception not be used during PEG- or Erwinia asparaginase treatment due to a potential interaction.

• Collection of serious adverse events was extended through long-term follow-up for compliance with clinical trial guidance. • A requirement for cytomegalovirus (CMV) polymerase chain reaction (PCR) testing at screening was added for participants with prior transplant. CMV PCR testing at screening was added to the schedule of assessments and analyte listing to clarify the required testing for these participants and those who are CMV positive, and to further clarify that CMV testing is recommended for all participants. • Clarification was added to the schedule of assessments that participants with symptoms of congestive heart failure during or after completion of induction therapy should have an echocardiogram before optional consolidation. • The timing of hematology collection in induction and consolidation for infants was corrected to day 36 (not day 35). • Clarification was added to the schedule of assessments the clinical chemistry parameters that are required beyond a standard metabolic panel. • A correction was made to remove cholesterol, HDL, LDL, and triglycerides from the analyte listing because they are not required for any exclusion criteria or monitoring. • Clarification was added that calculated creatinine clearance or glomerular filtration rate is required at screening. • Segmented neutrophils was added to the analyte listing to allow for calculation of total neutrophils. • The requirement for unstained slides of bone marrow aspirate and peripheral blood were removed.

• Clarification was added that chest imaging is only required for participants with a mass present at screening. • Pharmacodynamic blood sample collection was added at 60 minutes after the end of carfilzomib infusion on day 1 of induction to better monitor proteasome inhibition. • Pharmacodynamic and pharmacokinetic blood sample collection were added during consolidation to allow comparison between induction and consolidation • The dosing schedule of study treatments was separated from the schedule of assessments for consistency with phase 1 and other cooperative protocols. • Clarification was added that participants may be discontinued from study treatment due to death and for those determined to be ineligible for the study. • The objectives, endpoints, and analysis sets of the phase 2 portion were aligned with the statistical analysis plan (SAP). • Phase 2 endpoints were clarified (eg, removal of ‘full’ from complete response because it has no meaning). None of the edits change the endpoints themselves. • The treatment response definitions were updated per more recent United States Food and Drug Administration (US FDA) requirements and a definition for non-evaluable was added. • The definition of induction death was clarified with respect to the timing of response evaluation. • Clarification was added that assessment of response is central review of stained slides of bone marrow aspirate, peripheral blood, and minimal residual disease (MRD) by next generation sequencing (NGS) and local assessment of sites of extramedullary disease and MRD by flow cytometry and/or PCR or NGS if available • The missing baseline covariates imputation, sample size determination, and statistical analysis sections were updated to be consistent with SAP version 2.0, dated 19 November 2020 which was amended based on the feedback from the US FDA.

• Statistical comparison between experimental arm and external control arm of duration of response was removed and only descriptive statistics will be summarized because it will be analyzed only on responders in the Primary Analysis Set. • Subgroup analyses for participants who had anthracycline omitted from VXLD induction were added. Relapse time from prior transplant, number of prior relapses, and duration of most immediate prior remission subgroups were aligned with SAP. • Clarification was added that a sensitivity analysis will include any covariates still imbalanced after propensity score analysis. • The planned analyses for echocardiograms were added. • Details about the role of the independent biostatisticians who will support the futility efficacy analysis for Data Monitoring Committee (DMC) were added. • The investigators on the cover page were updated to reflect the anticipated involvement in the phase 2 portion of the study. • The anticipated number of sites participating in phase 2 was added. • A reference was added in the DMC section of the phase 1 protocol to the corresponding section for phase 2 to clarify the different roles of the DMC across phases. • Typographical and formatting issues were corrected throughout the document.

The protocol was amended as follow: • The recommended phase 2 dose of 20/56 mg/m2 carfilzomib identified from phase 1 by the Cohort Safety Review Committee was added. • Protocol-approved substitutions with dose equivalents were specified for components of VXLD or intrathecal chemotherapy for regions where these components are not available to allow enrollment globally while ensuring comparability of backbone therapy. • The omission of components of VXLD treatment (eg, anthracycline) under certain conditions or steroid from intrathecal chemotherapy is allowed to be consistent with current clinical practice or local standard of care. • For participants ≥ 12 months, the dose of daunorubicin during induction was changed from 25 mg/m2 for 4 doses to 60 mg/m2 and the dosing frequency was decreased from 4 times to once per cycle (reduced total dose from 100 to 60 mg/m2). This change is intended to reduce toxicity of the VXLD backbone by both reducing the total daunorubicin dose and to move it earlier in regimen to allow bone marrow function more time to recover. The change matches the design for daunorubicin used with the VPLD backbone in similar participants in COG Study AALL01P2 and Study AALL07P1 (VPLD + bortezomib). • The rationale for changes to the VXLD regimen in the phase 2 versus phase 1 portions of the study was added. • Vindesine was removed as an acceptable replacement for vincristine because no planned country/region is expected to require this substitution. • The discontinuation of asparaginase is allowed if Erwinia asparaginase is not available or desensitization procedure for asparaginase.

The protocol was amended as follow: • Remove the pharmacogenetic biomarkers from the exploratory objectives in phase 2. • Update inclusion criterion #114 to allow participants who received blinatumomab for treatment of MRD positive disease during first remission or for primary induction failure to achieve a first remission. • Update exclusion criterion #221 to require discussion with Amgen medical monitor for participants with less than 3 antibody half-lives since the last dose of monoclonal antibody. • Add exclusion criterion for known allergy to captisol in phase 2 of the study. • Add assessments for pancreatic function tests in the Schedule of Activities for phase 2. • Add assessments for echocardiograms in the Schedule of Activities for phase 2. • Update treatment of central nervous system-positive participants to allow alternatives to hydrocortisone substitutions per local standard practice. • Clarify instructions for lumbar puncture. • Clarify information collected during long-term follow-up. • Change day 29 to day 28 in the Schedule of Activities for dose regimen for consolidation therapy in participants aged less than 12 months at screening. • Update the reasons for early removal for protocol-required investigational product(s) or procedural assessments. • Update allowance for rescreening from 1 time to 3 times. • Add the use of asparaginase erwinia chrysanthemi recombinant (RYLAZE) as a replacement for polyethylene glycol asparaginase or Erwinia asparaginase in regions where approved and available for participants less than 12 months of age at screening receiving Berlin-Frankfurt-Munster consolidation therapy. • Administrative, typographical, and formatting changes were made throughout the protocol.

The protocol was amended as follow: • Allow the use of routine care procedures obtained within 7 days before enrollment to satisfy screening requirements. This is added due to a lack of clinical justification for repeating procedures that have no clinical benefit to the participant but expose the participant to potential risk. This flexibility is expected to reduce post-enrollment eligibility deviations and facilitate enrollment. • Update minimum enrollment numbers of participants with relapsed T-cell acute lymphoblastic leukemia (ALL) and with relapsed B-cell ALL based on current enrollment and health authority expectations. The minimum numbers of participants in the external control arm with relapsed B-cell ALL were also updated based on the status of incoming control arm data. The respective sample size estimates and expected power have been adjusted accordingly. • Respond to site feedback regarding confusion between phase 1 and phase 2 sections of the protocol. The synopsis is revised to remove phase 1 information and expand phase 2 information. Revisions to bring focus to phase 2 of the protocol result in a re-numbering of Sections 17 and beyond (eg, former Section 17 is now Section 18). Section and table names are revised to include “Phase 2” for clarity (throughout protocol). • Add details to define analytes used in the derivation of remission status for evaluation of the primary and secondary endpoints, and clarify that hematology and bone marrow aspirate/biopsy samples must be assessed locally • Revise the description of the primary endpoint to clarify that there is a single endpoint of CR. The previous amendment language included time window definitions for infants and children that could have been misinterpreted to suggest a CR endpoint for infants and a separate CR endpoint for children. The same clarification is made to the secondary endpoint for CR, CRp, CRh, and CRi after consolidation therapy.

• Allow bone marrow to be assessed by aspirate or biopsy or flow if morphology is not available, and permit bone marrows that are performed as part of routine medical care to be used for central laboratory minimal residual disease (MRD) assessment • In the schedule of activities, replace the line for adverse events, serious adverse events, and concomitant therapies review with “X” to prevent any confusion that these will be collected, at a minimum, from screening through safety follow up. No change is made to the timing of these activities. • Corrections to schedule of activities tables: added row for “other extramedullary disease assessment” and removed day 10 hematology, which was listed in error, from Table 20 (consolidation therapy for participants less than 12 months of age) • Remove exclusion criteria 212 and 213. These required a 3-month window after previous proteasome inhibitor therapy, and a 2-month window after previous VXLD (or similar). These exclusions were not present in the phase 1 portion of the protocol, their removal was requested by investigators, and is consistent with guidance to remove time-based washout periods from clinical trials eligibility (Harvey et al, 2021). • Correct inclusion criterion 113 so eligibility requires less than 5% blasts • Extend window allowed to make up missed doses from +1 day to +3 days • Allow alternate dosing or formulation of L-asparaginase in accordance with local standard of care • Allow substitution of intrathecal regimens based on local practice • Recommendation to collect medical history and surgical history 30 days prior to signing the ICF is removed to allow collection of full ALL medical history • Update treatment response definition in alignment with National Comprehensive Cancer Network guidelines (2022) and Shallis et al, 2021, consistency with Study 20180065, and Amgen endpoint guide for leukemia.

• Clarify details of the independent Data Monitoring Committee and independent Biostatistics Group, and remove requirement for every 3-month Steering Committee meetings • Update the covariates for subgroup analyses, and specify the categories into the SAP • To reduce confusion, the phase 1 pregnancy and lactation notification forms are removed with references to the respective forms to use for phase 2 • Safety reporting language updated per new Amgen protocol template updates • Revisions made for internal consistency of document (eg, number of allowed screen fails in Section 24.1.1 revised to align with Section 20.4; male contraception time revised in Section 22.1.6.15.3, Male participants to align with exclusion criteria) • Update investigator list on the cover page • Typographical and formatting issues were corrected throughout the document.

The protocol was amended as follow: • The approximate sizes of the external control arm datasets (Study 20180065) for the B-cell and T-cell primary analysis sets (B-PAS/T-PAS) were updated to approximately 74 B-cell acute lymphoblastic leukemia (ALL) participants and approximately 60 T-cell ALL participants from 90 and 70 participants, respectively. Previous versions of the protocol included external control sample sizes that were larger than required to support the analysis. • Language was added to the methods of interim analysis for futility to specify that the interim analysis will be done with the external control datasets that are available at time of interim. Previous versions of the protocol required the complete external control arm dataset to be available for interim analysis. • “Full” Analysis was revised to “Final” Analysis. Definition of Final Analysis Sets (T-FAS/B-FAS) was removed, as not required. A definition of a Final Analysis Set is not required due to the revisions to Interim and Primary Analyses. As revised, all external control data, as well as all 20140106 participants, will be included in the Primary Analysis (and in the Final Analysis). • Language was updated in the Primary and Interim Futility Analyses sections to align with the new approximate size of the external control arm datasets from Study 20180065. • Power estimates were adjusted based on the expected approximate number of external control arm datasets provided above. • Safety monitoring language was added in subsection 24.2.4.5.1, ‘Adverse Events of Special Interest’. • Additional changes were made for consistency and alignment throughout the protocol including, but not limited to, administrative, typographical, formatting, numbering, and abbreviation changes.