Low incidence of osteonecrosis in childhood acute lymphoblastic leukemia treated with ALL-97 and ALL-02 study of Japan Association of Childhood Leukemia Study Group

Osteonecrosis (ON) is a serious complication of ALL treatment and can affect quality of life on the long term. Incidence is variable and identified risk factors include older age, female sex, dose and type of corticosteroids, combination of dexamethasone and L-asparaginase, and ethnicity. This retrospective analysis of cohorts of Japanese patients with ALL examines the incidence, risk factors, and treatment of ON.
Incidence and characteristics of ON were determined in patients with ALL (Pre-B ALL or T-cell ALL) enrolled in 2 studies from the Japanese Association of Childhood Leukemia Study (JACLS) group (n=635 ages 1-15 years for ALL-97 and n=1,027 ages 1-18 years for ALL-02). In both studies, L-asp was administered with prednisolone and dexamethasone was not used.
24/1,662 patients had symptomatic ON during/after treatment. 15/24 were female. Most frequent site was the femoral head (17/24). 5-year cumulative incidence of ON in patients age 10 or older was 7.2% (ALL-97) and 5.9% (ALL-02), compared to most previous studies that report much higher incidence of ON, usually exceeding 10%. 11 patients required orthopedic surgery. In evaluating the risk factors, for both protocols the only significant risk factor was age greater than or equal to 10 years.
ALL-97 collected data prospectively while ALL-02 collected data retrospectively via questionnaire. Patients post BMT were excluded. Japan has a comparatively homogeneous population and thus their underlying genomic risk may be different from their genetically heterogeneous European or North American counterparts. Genomic data (such as SNP array data) might be helpful in untangling risk factors for ON.
Low frequency of ON in the JACLS studies compared to previous studies. Although the cumulative steroid dose and amount of dexamethasone was equivalent to previous COG/CCG/AIEOP trials, it was not co-administered with asparaginase so this may be an important risk factor. Genomic risk of ON was not assessed however and this may also be an important difference.

Acute Kidney Injury in Pediatric Patients Receiving Allogeneic Hematopoietic Cell Transplantation: Incidence, Risk Factors, and Outcomes

Acute kidney injury (AKI) is a common adverse event after hematopoietic cell transplantation (HCT). Although several small studies have evaluated AKI after pediatric HCT, large-scale studies are lacking. In addition, most earlier studies did not use standardized criteria of AKI.
This is a retrospective study, applied on a large pediatric population to establish the incidence and outcomes of AKI and to determine the risk factors associated with AKI after allogeneic HCT.
The study retrospectively analyzed data from 1057 pediatric and adolescent patients who received HCT from January 1991 to December 2015. Modified AKI network (AKIN) classification was used.
The 100-day cumulative incidences of all stages of AKI, stage 3 AKI, and AKI requiring renal replacement therapy (RRT) were 68.2%, 25.0%, and 7.6%, respectively. Overall survival at 1 year was not different between patients without AKI and those with stage 1 or 2 AKI but was significantly different between patients without AKI and patients with stage 3 AKI with or without RRT requirement (66.1% versus 47.3% versus 7.5%, respectively; P.001). Age, year of transplantation, donor type, SOS, and acute GVHD were independent risk factors for stages 1 through 3 AKI. Age, donor, conditioning regimen, number of HCTs, SOS, and acute GVHD were independent risk factors for AKI requiring RRT.
This study does provide a large-scale and comprehensive analysis of AKI from HCT yet with some limiting factors.
First, relying on serum creatinine as a sole measure of AKI with no data on urine output or state of hydration carries potential flaws in the applicability of the modified AKIN criteria. As in these categories of patients receiving HCT for prolonged, complex illnesses, serum creatinine may remain low despite a marked reduction in glomerular filtration rate. Serum creatinine is influenced by multiple factors including diet and muscle mass.
Second, patient enrolment age up to 27 years including young adults might hamper conclusions for pediatric patients.
Third, including patients who received multiple transplants, yet their study enrolment renal condition was not elucidated.
AKI was a prevalent adverse event affecting more than two-thirds of pediatric recipients receiving allogeneic HCTs. Notably, 25% of patients experienced severe stage 3 AKI according to the AKIN criteria, which greatly affected survival outcomes.

Measurable residual disease detection by high-throughput sequencing improved risk-stratification for pediatric B-ALL.

The initial response to therapy is a key prognostic factor in pediatric B-ALL. This is currently determined as the level of measurable residual disease (MRD) at the end of Induction chemotherapy. Normally determined by flow cytometry (FC), this leads to limitations in the analytic sensitivity and difficulty in clinical standardization of the technique. As well, changes in antigen expression on the lymphoblasts during therapy may obscure detection by FC. High-throughput sequencing (HTS) of the immunoglobulin and T-cell receptor loci may represent a superior methodology for detection of MRD.
Paired pre-treatment and end of induction samples for patients enrolled in front-line childhood B-ALL treatment trials were identified for MRD evaluation. Both FC and HTS were performed on the samples. For HTS, pretreatment samples were assessed for clonality by deep sequencing of the immunoglobulin (IGH@) gene. A dominant sequence determined for identification at end of induction, as an indicator of MRD. The amount of DNA was extrapolated to the equivalent total nucleated cell count, to give a comparable metric to MRD by FC.
607 evaluable samples were studied, and HTC and FC showed good correlation for both event-free and overall survival at an MRD of 0.01%. Samples where MRD was above the threshold by HTS but not FC were found to have an intermediate prognosis, suggesting that it had a higher analytic sensitivity for low levels of MRD. Patients with no trackable rearrangement of the immunoglobulin heavy chain by HTS, suggesting an earlier stage neoplastic transformation, had a poorer prognosis.
HTS required initial access to heavy leukemic involvement to identify an appropriate target sequence, and cannot be used in patients lacking a detectable immunoglobulin or T-cell receptor rearrangement. Turn-around time is also longer than with FC and the technology is currently more expensive.
HTS offers an improved analytic platform for measurement of MRD in pediatric B-ALL, with the same threshold as standard FC.

Early Nutrition Intervention Attenuates Weight Gain for Pediatric Acute Lymphoblastic Leukemia Patients in Maintenance Therapy

Children with ALL are at increased risk for long-term obesity. Recent studies have indicated that weight gain begins during ALL therapy, particularly during induction and maintenance phases, and often persists beyond treatment. There are no published strategies specific to the pediatric ALL population to reduce obesity.
In a retrospective manner, this study investigated the impact of a 3-visit nutrition intervention early in maintenance therapy involving 3 one-on-one visits with a registered dietician. The visits focused on the child's current nutritional status and goals and provided handouts. 33 patients who received the intervention were compared with a control group of 34 historical patients. BMI was recorded at diagnosis, end of induction, beginning of delayed intensification, day 1 of maintenance therapy and monthly for 12 months.
As expected, the mean BMI z-score increased from diagnosis to day 29 of induction, dropped in delayed intensification and then rose throughout maintenance therapy. There was no statistically significant difference between the BMI z-score of the control group and the intervention group at the different time points. On multivariate analysis, the intervention group's BMI z-scores increased less over maintenance therapy compared to the control group (P=0.0001). However, the intervention was less effective for patients who gained more in BMI z-scores between diagnosis and the start of maintenance therapy.
Information on behaviours including diet and exercise and other indicators of body composition including fat composition were not collected as it was a retrospective chart study. The patients who received the intervention was at the discretion of their health care team. The small sample size, potential selection bias, and use of a historical control severely limits the generalizability of the results.
This study does not provide evidence for efficacy of this particular intervention but does not mean that educational interventions are not useful. Future prospective trials should be well designed and draw from the extensive experience in the adult literature.

Integrated molecular profiling of juvenile myelomonocytic leukemia

JMML is a rare myeloid neoplasm with few therapeutic options. The majority of cases are driven by mutations in the RAS pathway and several molecular and clinical prognostic markers are known. This research group used multi-omic profiling to find driver events in the ~20% of patients without RAS mutations and to identify new prognostic markers.
150 children were included in this study although not all patients were evaluated by each modality. The majority of data was generated through high-throughput exome sequencing, high-throughput mRNA sequencing, and a DNA methylation array (the now-defunct Illumina 450K array).
1) Three children without RAS pathway mutations were found to have ALK (n = 2) or ROS1 (n = 1) fusions. Crizotinib (a TKI effective against ALK and ROS1 activity) decreased proliferation in cell lines obtained from these patients. One patient with an ALK fusion and refractory disease had a clinical response to Crizotinib bridging her to transplant.
2) The group identified 2 DNA methylation profiles and 4 RNA expression profiles. LIN28B (a developmental gene involved in microRNA activity and affected in multiple cancers) was both differentially methylated and expressed. The "hypermethylation" profile and "AML-like" profile both carried poor prognosis.
When creating tumor profiles using genome-wide data such as DNA methylation or RNA sequencing, investigators must make many choices (e.g. how to normalize and filter the data, which coordinates to use, how to define a subgroup etc). In this study, the selected sequencing methodology is debatable, so analyses can only be considered robust if they are convincingly replicated - otherwise they may be artifacts of upstream analytical decisions
This is a thorough analysis of a large number of JMML cases. The finding of ALK/ROS1 fusions - even though rare - should prompt us to look for these fusions in our refractory patients keeping in mind the possible use of targeted agents. The DNA methylation and expression profiles require replication before integration with clinical decision making.

Value of flow cytometric analysis of peripheral blood samples in children diagnosed with acute lymphoblastic leukemia

Peripheral blood samples are frequently screened by flow cytometry before bone marrow for suspected leukemia to facilitate treatment decisions. Criteria to establish the diagnosis of a lymphoblastic malignancy from peripheral blood are not well defined.
Retrospective comparison of paired results of peripheral blood flow cytometry and bone marrow in 383 children with ALL diagnosed consecutively at a single center from January 2007 to February 2016. Patients were aged 0-18 years and had an adequate peripheral blood sample collected up to 7 days prior to the BMA. Four-color flow cytometry was used until September 22, 2014, then a 10-color panel was established thereafter.
Of 383 patients with B or T precursor ALL and paired results, only 3 patients had discordant results. There were 2 false positives peripheral blood samples (corresponding to lymphoblastic lymphoma with BM involvement below the threshold of leukemia and ALL that initially did not meet diagnostic criteria but later progressed) and one false negative (qualitatively positive upon review). In 75% of patients (289/383) who underwent both peripheral blood flow cytometry and BMA, the diagnostic LP and first dose of IT chemo were performed during the same sedation as the BMA.
Hematopathologist experience may lead to bias.
High concordance of results between peripheral blood and bone marrow flow cytometry in diagnosis of ALL. 1% or more blasts in the peripheral blood anticipated a diagnosis of ALL in the subsequent BMA with high sensitivity and specificity. Peripheral blood should not replace bone marrow in the diagnosis of ALL but integration into diagnostic approach could help with scheduling initial procedures, reduce anesthesia procedures, and optimize use of healthcare resources.

Brentuximab vedotin plus bendamustine in relapsed or refractory Hodgkin’s lymphoma: an international, multicentre, single-arm, phase 1–2 trial

The objective of this study was to explore the safety and clinical activity of brentuximab vedotin plus bendamustine in heavily pretreated patients with relapsed or refractory Hodgkin's lymphoma and anaplastic large T-cell lymphoma.
This was an international, multicentre, single-arm, phase 1-2 trial. Patients with Hodgkin's lymphoma or anaplastic large-T-cell lymphoma who had received at least one previous multi-agent chemotherapy regimen.
The trial occurred in 2 phases:
Phase 1: patients were assigned following a 3+3 dose-escalation design to one of four cohorts to receive brentuximab on day 1 of a 21-day cycle and bendamustine on days 1 and 2 of a 21-day cycle. Outcomes were maximum tolerated dose and dose limiting toxicity.
Phase 2: all patients received brentuximab plus bendamustine at the recommended dose from phase 1.
Complete response was defined as PET negativity.
65 patients were enrolled, most with Hodgkin's lymphoma (n=64). The age range of patients was 18-72 years. 28 patients were evaluated in the phase 1 study and 37 patients were evaluated in the phase 2 study. Maximum tolerated doses were not reached. Dose limiting toxicities included grade 4 neutropenia and diffuse rash. The dose used for phase 2 was the standard dose of each agent when used independently. An overall response was achieved in 29/37 (78%) of patients. Complete response was obtained in 16/37 patients (43%).
6 patients in the phase 1 study and 3 patients in the phase 2 study had progression of disease during treatment (9/65, 14%).
This was a non-randomized trial in a heavily pre-treated group of patients. There are only few patients included in this study. Only adults participated in this study with a range of ages from 18-72 years.
Brentuximab vedotin and bendamustine achieved a clinical response in relapsed and refractory adult patients with Hodgkin's disease.

Young Female Donors Do Not Increase the Risk of Graft-versus-Host Disease or Impact Overall Outcomes in Pediatric HLA-Matched Sibling Hematopoietic Stem Cell Transplantation.

This is a retrospective cohort study, in 244 pediatric patients, to address the hypothesis that the presence of T and B cells sensitized by exposures during pregnancy are a contributing factor to differences in outcomes between sex-matched and sex-mismatched transplants. Theoretically, there should be no such difference when a non-exposed (non-alloimmunized) female donor is used. In the study, they assume that young (12 years) female donors are a sexually naïve population and therefore the presence of alloimmunization including to H-Y antigens should be minimal.
Data from 244 pediatrics patients were analyzed. The outcome was the development of acute grade II to IV GVHD and Chronic GVHD. Survival analysis was assessed at 100 days, 1 year and 5 years. Age was dichotomized to improve the interpretability of the results.
Donor age >12 yrs represents 50% of the population in the study. Univariate analysis revealed older patient age, older donor age, conditioning with CY-TBI and earlier year of transplant as significant predictor of aGVHD. Of these, all but donor age showed significance in multivariate analysis.
The effect of female donor sex on cGVHD noted in the model adjusted for patient age, HLA match, and stem cell source lost significance if the donor was 12 years old, but increased in magnitude and was significant if the donor was ≥12 years old (OR, 13.6; 95% CI, 2.8 to 39.6). Patient age was not a significant risk factor in multivariate analyses.
Population sample inclusion criteria did not consider important factors as:
donor history of blood transfusions and the wide age range of participants. Age was used as a surrogate for sexual-naivety and therefore it does not directly answer the question.
The study concluded that when selecting among sibling donors for a pediatric patient, priority should be given to donors 12 years of age or younger and that selection can be done independently of donor gender and sex match. Other explanations for the results of this study cannot be excluded and many limitations on patient enrolment criteria, age, and confounding factors could have impacted the results.

Oncogenic mutations combined with MRD improve outcome prediction in pediatric T-cell ALL

Risk stratification for pediatric T-ALL is based primarily on clinical findings and MRD. This study aimed to identify new genetic prognostic factors to improve the detection of patients at risk of relapse. Mutations in the Notch1 and Ras pathway were selected based on their recognition as oncogenic pathways in T-ALL and the reports in adult literature demonstrating their prognostic value.
220 patients treated prospectively on the FRALLE 2000T study (France) from 2000 to 2010 who had DNA material available were retrospectively analyzed for mutations in Notch1 (N), FBXW7 (F), K-RAS/N-RAS (R) and PTEN (P), both somatic and germline. The low-risk group was defined as having N/F mutations in the absence of R/P mutations. High-risk group was defined as having both N/F and R/P somatic mutations or either N/F germline or R/P germline mutations. Multi-variable regression analyses were performed to assess whether this classification of patients into low (n=111) and high-risk (n=109) groups was prognostic. Of note, on the FRALLE 2000T study, risk stratification was based on MRD.
Classification into low-risk and high-risk genetic groups was an independent prognostic risk factor. 5-year cumulative incidences of relapse were 11% and 36% for low-risk and high-risk groups respectively. When combined with WBC count and MRD, genetic risk groups helped further identify patients at low and high risk of relapse: Patients with a WBC > 200,000/µL, high-risk genetics and MRD > 1x10e4 had a cumulative incidence of relapse of 46% as compared to only 2% for patients with a WBC 200,000/µL, low-risk genetics and MRD 1x10e4.
Retrospective analysis of a prospective cohort. Similar study based on UK2003 study did not demonstrate significance of K/N-ras and PTEN mutations.
The combination of white cell count > 200,000/µL, MRD positivity, and genetic risk group were helpful in stratifying relapse risk in pediatric T-ALL, in particular identifying the low-risk group, in this cohort. These findings are in contrast with the findings from the UKALL2003 data which did not find genetic risk factors contributory (https://www.ncbi.nlm.nih.gov/pubmed/26220040).

IKZF1plus Defines a New Minimal Residual Disease-Dependent Very-Poor Prognostic Profile in Pediatric B-Cell Precursor Acute Lymphoblastic Leukemia.

This paper from the International BFM Study Group further defines the IKZF1 gene as a very poor prognostic marker in pediatric B-cell precursor ALL. The study group looked to refine the prognostic strength of the IKZF1 deletion by looking at the effect of the co-occurring gene deletions.
The study analysed 991 patients with B-cell precursor ALL from the European group (AIEOP-BFM) trial with complete information for copy number alterations of major genes associated with outcome in ALL. There was also a smaller replication cohort of 417 patients from the same trial.
The study then analysed gene combinations and evaluated patient outcomes including how gene combinations affected outcomes combined with MRD status.
"IKZF1plus" was defined as an IKZF1 deletion co-occurring with deletions in CDKN2A, CDKN2B, PAX5 or PAR1 in the absence of a deletion in the ERG gene.
The IKZF1plus group made up of 6% of patients in the cohort of children with B-cell precursor ALL (n=63). The 5-year EFS for this IKZF1plus group was 53% compared to 79% for those with lone IKZF1 deletions and 87% for patients who lacked the IKZF1 deletion altogether. The 5-year EFS when combining the IKZF1plus combination with MRD was 95% for standard risk MRD versus 40% for intermediate MRD and 30% for high-risk MRD.
Given differences in therapy and timing/method of MRD measurement it's not clear whether these results can be generalized to other study groups.
The use of the IKZF1 deletion in combination with specific additional single gene deletions – the so-called IKZF1plus – provides an independent and strong molecular stratification marker in addition to MRD measurements. The IKZF1plus group with positive MRD are a particularly high-risk group and should be assigned additional/experimental treatments, which will be evaluated in the upcoming AIEOP-BFM ALL 2017 trial. On the other hand, the IKZF1plus genotype loses prognostic significance when MRD is negative.

Genotype-Specific Minimal Residual Disease Interpretation Improves Stratification in Pediatric Acute Lymphoblastic Leukemia

Cytogenetic analysis and minimal residual disease (MRD) assessment are used to personalize treatment in current protocols for acute lymphoblastic leukemia (ALL). MRD status is currently assigned using cut-off limits in treatment protocols. This study assessed MRD as a continuous variable using PCR-based MRD assessment (Ig/TCR rearrangements) and correlated results with somatic genetic changes.
3,113 consecutive patients with ALL treated in the MRC UKALL2003 protocol (2003 to 2011) were eligible and 2,542 were analysed. Patients were classified into four mutually exclusive cytogenetic genetic groups: good risk: ETV6-RUNX1, high hyperdiploidy (51 to 65 chromosomes); high risk: KMT2A (MLL) fusions, near haploidy, low hypodiploidy ( 40 chromosomes), iAMP21, and TCF3-HLF; intermediate risk: TCF3-PBX1 and all other patient-cases with B-ALL; and patients with T-ALL. MRD at end of induction was assessed as a continuous variable with a minimum detection level of 1x10e-5.
MRD results were highly dependent on genetic risk groups with ETV6-RUNX1 but also TCF3-PBX1 showing rapid MRD-clearance (in 36% and 43% respectively). Patients with iAMP21 on the other hand had a high rate of recurrence even in those with negative MRDs. T-ALL patients more frequently had not reportable results which renders the Ig/TCR PCR-based technique less reliable for this patient group. In T-ALL, outcomes were associated with MRD-negativity or frank positivity (>=5%) but not intermediate results.
This is a post hoc analysis of MRD results and treatment was not based on these findings. The impact of continuous MRD assessment was therefore not assessed on possible treatment modifications.
As previously known, PCR-based high-sensitive MRD-assessment is a powerful tool and results differ between different cytogenetic groups. The inclusion of iAMP21 as a high-risk somatic change per se was supported.

Reduced-Intensity Delayed Intensification in Standard-Risk Pediatric Acute Lymphoblastic Leukemia Defined by Undetectable Minimal Residual Disease: Results of an International Randomized Trial (AIEOP-BFM ALL 2000)

Delayed intensification (DI) in acute lymphoblastic leukemia (ALL) is an essential element in treatment protocols but is intensive and associated with toxicity. This study originated from the AIEOP-BFM 2000 trial and aimed at testing non-inferiority of reduction in DI from 49 days (P-II) to 29 days (P-III) with a shorter duration of steroids and lower doses of cyclophosphamide, vincristine, and doxorubicin.
The randomized prospective trial AIEOP-BFM ALL 2000 is a European multinational cohort study. From 2000 to 2006, 4,937 patients were enrolled on the trial. Of 1,346 patients with standard risk criteria (SR, the lowest risk category), 1,164 were randomly assigned on the reduced regimen P-III or the standard regimen P-II (roughly half in each arm).
Median follow-up time was 8.4 years. The disease-free survival was 91.8% vs. 95.8% with the reduced regimen vs. the standard treatment with 62 versus 42 events (p=0.04). 8-year OS was 96.1% versus 98% (NS). Acute toxicity was about the same in the two regimens with slightly more life-threatening events occurring with standard therapy P-II (n=10 vs. 7). The reduced regimen P-III was not associated with similar outcomes compared to the standard regimen P-II and therefore, treatment reduction in SR patients with ALL was not successful in this trial. The only subgroups with similar outcomes were ETV6-RUNX1 positive ALL and patients aged 1 to 6 years.
The results of this trial are specific to the BFM backbone and risk stratification which differs compared to other large international trials.
Treatment reduction in delayed intensification was not successful in this trial for Standard risk ALL patients. ETV6-RUNX1 status was added as a favorable cytogenetic marker in the AIEOP-BFM 2009 trial. The 2009 trial will ask whether a reduction of the 4-drug induction by using half the daunorubicin dose in low-risk patients leads to comparable outcomes.