Poverty, race, ethnicity, and survival among U.S. children with non-metastatic osteosarcoma treated on EURAMOS-1: A report from the Children’s Oncology Group

10004

Background: Children living in poverty and those who identify as a race/ethnicity other than non-Hispanic White experience higher rates of relapse and lower overall survival across many pediatric cancers. Racial, ethnic and socioeconomic disparities have not been comprehensively investigated in osteosarcoma. We leveraged data from US-enrolled patients on the recent international EURAMOS-1 trial to investigate disparities in survival outcomes. Aim: Identify if race/ethnicity, household or neighborhood poverty exposure are associated with event-free survival (EFS) or overall survival (OS) in patients with non-metastatic osteosarcoma enrolled at a US-center on EURAMOS-1. Methods: Retrospective cohort study of US patients aged 5-21 years enrolled on EURAMOS-1 with a diagnosis of non-metastatic, primary osteosarcoma. Poverty was the primary exposure defined at the household- (sole coverage with Medicaid or CHIP public insurance versus other) and neighborhood- (Census-defined high-poverty ZIP code with >20% of residents living at <100% Federal Poverty Level vs other) levels. Race and ethnicity were categorized to reflect structural inequities and historically marginalized populations, as Hispanic, non-Hispanic Black (NHB), non-Hispanic Other (NHO), and non-Hispanic White (NHW). OS and EFS as a function of time from trial enrollment were estimated using the Kaplan-Meier method. Hypotheses of associations between risks for EFS-event, death and post-relapse death with poverty-exposures and race/ethnicity were assessed using log-rank tests. Statistical comparisons were performed excluding patients with missing values for the exposures considered. P-values <=0.05 were considered significant. Results: Among 758 patients, 27% were household poverty-exposed and 29% were neighborhood-poverty exposed. Twenty-one percent of children identified as Hispanic, 17% NHB, 5% NHO and 57% NHW. Neither household- or neighborhood-poverty, nor race/ethnicity were significantly associated with risks for EFS-event or death. Post-relapse risk for death differed significantly across race/ethnicity with NHB at greatest risk compared to others (4-Year post-relapse survival 35.7% Hispanic vs. 13.0% NHB vs. 43.8% NHO vs 38.9% NHW; p=0.0046). Conclusions: Neither poverty-exposures nor race/ethnicity were associated with EFS or OS in this COG trial-enrolled cohort, suggesting equitable outcomes following standardized therapy. Despite this, non-Hispanic Black children experienced significantly inferior post-relapse survival. Investigation of mechanisms driving these post-relapse disparities, including inequities in health care delivery and access to post-relapse trial-enrollment, are paramount to ensure equity in outcomes for all children with cancer.

Long-term outcomes in patients with localized Ewing sarcoma treated with interval-compressed chemotherapy: A long-term follow-up report from Children’s Oncology Group study AEWS0031

11505

Background: Children’s Oncology Group study AEWS0031 demonstrated superior 5-year event-free survival (EFS) in patients with localized Ewing sarcoma (ES) receiving interval-compressed (IC) chemotherapy (every 2 weeks) compared to standard timing (ST) chemotherapy (every 3 weeks). We assessed the long-term outcome of patients treated on AEWS0031 to determine whether the survival advantage of IC chemotherapy was maintained at 10 years. Methods: AEWS0031 enrolled 568 eligible patients with localized ES. Patients were stratified into four groups by age (<18 years and ≥ 18 years) and primary site (pelvic and non-pelvic), and randomized to receive 14 cycles of alternating vincristine-doxorubicin-cyclophosphamide and ifosfamide-etoposide given every 3 weeks (ST; Regimen A) vs. every 2 weeks (IC; Regimen B). For this updated report, one patient was excluded due to uncertainty of original diagnosis giving a total of 567 patients in this analysis. Data for tumor measurements and histologic response were collected retrospectively from institutional reports. EFS and overall survival (OS) were estimated using the Kaplan-Meier method and compared using the log-rank test and Gray’s test for cumulative incidence (CI). Results: The 10-year EFS for patients treated with IC chemo was 70% compared to 61% for ST chemo (p = 0.03), and the OS was 76% with IC chemo compared to 69% for ST chemo (p = 0.03). The 10-year CI of second malignant neoplasms (SMNs) for ST chemo was 4.2% [95% confidence interval: 2.4-7.5] compared to 3.2% (95% confidence interval: 1.6-6.3) for IC chemo (p = 0.5). There was a trend towards improved 10-year EFS in those receiving IC chemo both with non-pelvic (N = 477; 71% vs. 64%) and pelvic (N = 90; 67% vs. 43%) primary tumors. Similarly, the 10-year EFS was superior for patients treated with IC chemo in both the < 18 years (N = 500; 73% vs. 64%) and ≥ 18 years (N = 67; 53% vs. 37%) age groups. Among the 184 patients with available histologic response data, the 10-year EFS from the time of local control was 76% for those with < 10% viable tumor and 56% for those with ≥ 10% viable tumor (p = 0.01). Additional analysis comparing patients with any viable tumor vs. no viable tumor (NVT) by treatment regimen demonstrated that patients with NVT who received IC chemo had 10-year EFS and OS from local control of 91% and 97%, respectively. In the 210 patients for whom tumor volume calculations were possible, there was no difference in the 10-year EFS for patients with tumors < 200 mL vs. ≥ 200 mL. Conclusions: With longer term follow-up, IC chemotherapy for localized ES is associated with superior EFS and OS without an increase in SMNs. This study suggests patients ≥ 18 years with poor necrosis or pelvic primary tumors remain at high risk for relapse despite IC chemo, emphasizing the need for alternative treatment strategies to improve their outcomes. Clinical trial information: NCT00006734.

A phase II multi-arm study to test the efficacy of oleclumab and durvalumab in specific sarcoma subtypes

TPS11594

Background: Anti-PD 1/PD-L1 blockade alone or in combination with anti-CTLA4 have yielded suboptimal results in most sarcoma subtypes. CD73, an ectonucleotidase, catalyzes the rate-limiting step for adenosine production in the extracellular space, which then aids tumors in evading immune recognition and destruction. Oleclumab, a monoclonal antibody (mAb) selectively binds and inhibits the activity of CD73, and preclinical data suggests additive activity with durvalumab, a mAb that blocks PD-L1. We designed a trial combining oleclumab and durvalumab in certain sarcoma subtypes, selected based on modest activity with anti-PDL-1 and intense staining with CD73 in the tumor microenvironment. Methods: This phase 2 study (NCT04668300) is enrolling patients with age ≥18 years with recurrent/metastatic angiosarcoma (cohort 1) or dedifferentiated liposarcoma (DLPS) (cohort 2) and ≥12 years with recurrent/metastatic osteosarcoma (cohort 3), who have received at least one prior systemic therapy but are checkpoint inhibitor naïve and have measurable disease. Each treatment cycle is 28 days with oleclumab administered at 3000 mg i.v. every 2 weeks x 5 doses, and then every 4 weeks and durvalumab administered at 1500 mg i.v every 4 weeks. Tumor assessments are based on RECIST 1.1 and immune-related response criteria (irRC) and performed at baseline, and every 8 weeks after start of therapy, with an additional scan at 12 weeks for confirmation of response. Planned sample size is ≤ 25 pts in each arm. The primary efficacy endpoint for cohorts 1 and 2 is response rate (RR) at 4 months (per RECIST 1.1). The primary efficacy endpoint for cohort 3 is event free survival (EFS) rate at 4 months. If there is a high probability that the RR4 months is unlikely to be at least 20% for cohorts 1 and 2 or the EFS4 months is unlikely to be at least 40% for cohort 3 then the accrual of the corresponding cohort will be halted. The cohorts will be monitored separately for both futility and toxicity in groups of 5 after a minimum of 10 patients have been enrolled in each cohort. Secondary endpoints for the study include safety, best RR by RECIST and irRC, median PFS, and OS. Core needle biopsies and blood samples are collected at baseline and early on-treatment (week 6). Fresh flow cytometry is being performed to assess changes in T-cell activation, proliferation, and function and CD73 expression in the membrane and cytoplasm is being assessed by immunohistochemistry. Localization of tumor-infiltrating lymphocytes and engagement of the PD-1/PDL1 axis is being assessed using multiplex immunofluorescence staining. As of Jan 30th, 2022, twenty-two patients have initiated study treatment, 3 in cohort 1, 12 in cohort 2, and 7 in cohort 3. Clinical trial information: NCT04668300.

Clinical characteristics and outcomes of adult alveolar rhabdomyosarcoma (ARMS) patients on front-line therapies: An MD Anderson Cancer Center (MDACC) case series

e23548

Background: Rhabdomyosarcomas are the most common soft tissue sarcoma in children, and the prognosis of pediatric ARMS has improved with the use of multi-modality therapies. However, ARMS in adults is rare, and long-term outcomes continue to be poor. This study aimed to evaluate clinical outcomes of an adult ARMS population on different front-line systemic chemotherapies, particularly the vincristine/doxorubicin/ifosfamide (VDI) regimen. Methods: Adult fusion-positive confirmed ARMS patients over the age of 18 years (y) treated at MDACC from 2004 to 2018 were identified in our patient registry. Descriptive clinical statistics including staging, front-line chemotherapy, multimodal therapy usage, and survival analyses were performed. Results: 49 patients were identified, with mean age of 34.9 y (range 18y - 67y), and 53% were male. Most patients were white (53%, 26 pts), and the most common primary tumor site was the parameningeal space (63%; 31 pts). Patients were either intermediate (67%) or high clinical risk (33%). Most patients were IRSG clinical group IIIa (36%), IIIb (20%) or IV (33%) and were classified clinical stage 3 (49%) or 4 (33%). Of all patients at diagnosis, 71% had nodal disease and 32% were metastatic. Radiotherapy and surgery were given with upfront chemotherapy in 33 pts (67%) and 24 pts (49%) respectively, with 19 patients receiving both. Median OS for the entire cohort was 3.6 years. Doxorubicin containing chemotherapy regimens trended to worse OS than non-doxorubicin containing regimens (2.3 yrs vs 4.0 yrs, p = 0.355). Comparing patients who received VDI (19 pts) vs non-VDI (30 pts; 13 received Actinomycin D, 12 received doxorubicin in different regimens, and 5 received neither), median OS was 1.8 yrs vs 3.8 yrs (p = 0.283) respectively. There were similar number of front-line chemotherapy cycles (8.5 vs 9.5 cycles), high clinical risk (26% vs 37%) and metastatic disease (21% vs 36%) in the VDI vs non-VDI cohorts. Patients receiving upfront radiation had improved survival (3.7 vs 1.5 yr, p = 0.01), but this is likely confounded by those with metastases being less likely to receive upfront radiation. Conclusions: In this single center retrospective analysis of adult ARMS patients, survival outcomes continue to be similar to historical outcomes. There was no statistically significant OS difference in patients who did or did not receive doxorubicin containing front-line chemotherapy regimens, or in particular VDI therapy, although there was a trend to decreased OS. However, limitations to this study include limited sample size, non-randomization to treatment selection, and possible biases in patient selection for different chemo regimens. Based on these observations, randomized prospective studies are necessary to delineate which frontline chemotherapy regimen is most beneficial in this rare tumor in adults.

Testing of B7-H3 targeting antibody-drug conjugate (ADC) MGC018 in models of pediatric solid tumors by the Pediatric Preclinical Testing Consortium (PPTC)

10037

Background: B7-H3 (encoded by the CD276 gene) is an immunoregulatory molecule that is widely expressed in pediatric embryonal tumors and sarcomas, while expression is limited for normal tissues. Among PPTC models, osteosarcoma models showed highest CD276 transcript levels followed by Wilms tumor, neuroblastoma, rhabdomyosarcoma, and Ewing sarcoma. Protein expression by IHC generally followed expression at the RNA level, with high expression by IHC observed across a range of models. MGC018 is a duocarmycin-based humanized ADC targeting B7-H3 that shows selective cytotoxicity for B7-H3 expressing cancer cells, robust in vivo activity against a range of adult cancer preclinical models, a favorable pharmacokinetic and safety profile in cynomolgus monkeys, and has entered clinical testing for adults with cancer. Herein we report the in vivo antitumor activity of MGC018 against preclinical models of pediatric solid tumors. Methods: MGC018 was tested at 6 mg/kg administered intraperitoneally (IP) on Day 1. SYD988, an anti-CD20 ADC with the same linker and payload as MGC018, was used as a control arm at 6 mg/kg IP on Day 1. Osteosarcoma models, due to their slower growth kinetics and lower rates of tumor regression, were tested with n = 10 mice per arm, while the sarcoma and neuroblastoma models were tested with n of 1 or n of 2 designs, respectively. Activity was evaluated using response criteria mirroring clinical criteria (progressive disease (PD), stable disease (SD), partial response (PR), complete response (CR), and maintained CR (MCR)], by minimum relative tumor volume across all models, and by event-free survival (EFS) comparisons between treatment groups. Results: For neuroblastoma models (n = 9), 3 and 2 neuroblastoma models showed objective responses (PR/CR/MCR) or SD, respectively, to MGC018, while none showed an objective response or SD to SYD988. For osteosarcoma, results are mature for 1 of 5 models, with this model showing an MCR response to MGC018 and a PD response to SYD988. For Ewing sarcoma (n = 5), a PR and MCR were observed to MGC018 with the remaining models showing PD; for SYD988 a single CR was observed with the remaining 4 models showing PD. For rhabdomyosarcoma (n = 3) 2 MCR and 1 PD were noted for MGC018, while only PD was noted for SYD988. EFS duration exceeding 100 days to a single dose of MGC018 was observed for 2 rhabdomyosarcoma, 1 Ewing sarcoma, and 1 neuroblastoma model. Evaluation of the relationship between B7-H3 expression and response to MGC018 will be performed when results from all studies are completed. Conclusions: B7-H3 is an important target for immuno-oncology agents for childhood cancers. Our results provide proof-of-principle for the ability of MGC018 to produce profound antitumor activity in select pediatric solid tumor models in a B7-H3 specific manner.

Outcome of patients with recurrent/refractory osteosarcoma enrolled in three recent phase II trials: A report from the Children’s Oncology Group

11530

Background: Based on seven prior Phase 2 Children’s Oncology Group (COG) trials, the 4-month event-free survival(EFS) and associated confidence interval for relapsed osteosarcoma with measurable disease according to RECIST was determined to be 12% (95% CI 6 - 19%). Three prospective clinical trials were conducted using this historical benchmark to detect activity defined by an EFS improvement of double the upper confidence interval. This report summarizes the outcome of these studies, describes whether the historical data remains an accurate baseline, and considers implications for future phase II trial study design in relapsed osteosarcoma measurable according to RECIST. Methods: We conducted an analysis of outcome for patients with recurrent/refractory osteosarcoma enrolled in three recent prospective COG phase II trials; AOST1321 (unresected cohort), AOST1322 and AOST1521 that used EFS at 4 months as the primary endpoint. Patients were eligible if they had osteosarcoma that had recurred or become refractory after standard therapy and had measurable disease according to RECIST. We assessed whether risk of an EFS event is modified by age, sex, race/ethnicity, number of prior chemotherapy regimens, or time to first relapse. Results: In each of the three phase II trials (unresected group of AOST1321, AOST1322, AOST1521), the drugs tested (denosumab, eribulin and glembatumumab) were concluded to be not effective due to a failure of the patient populations to meet the prespecified active 4-month progression free survival endpoint. The 4-month EFS for the 57 evaluable patients enrolled on these trials was 7% (95% CI 2 – 16%), similar to the 4-month EFS for 96 patients in the previous seven phase II trials of 12% (95% CI 6 - 19%). The combined EFS at 4 months for all 10 studies is 10% (95% CI 6 – 15%).There was no significant difference in EFS across trials based on age, sex, ethnicity, number of prior treatment regimens, consistent with prior analysis. Data from AOST1321 and AOST1521 were analyzed to determine the impact of time to first recurrence on EFS. Two different quantifications were applied: 1 year or less versus 2 or more years; and 2 years or less versus 3 years or more. Neither categorization was statistically significant. Conclusions: The EFS at 4 months in the three new phase II trials is similar to the previous seven phase II single arm trials. The combined analysis of 153 patients from 10 trials tightens the confidence interval, moving the upper 95% CI to 15%. Modification to future study designs could be considered based on this updated analysis. EFS at 4 months remains a robust primary endpoint. Single-arm trials using this endpoint based on the historical benchmark have accrued rapidly and allowed assessment of multiple novel agents in osteosarcoma. The negative trial results and continued poor outcome highlight the need for new approaches for relapsed osteosarcoma. Clinical trial information: NCT02097238, NCT02470091, NCT02487979.

Cell-surface vimentin positive circulating tumor cells as a diagnostic and prognostic indicator in pediatric sarcoma patients

e15022

Background: Despite advances in care, the 5 year overall survival for patients with relapsed and or metastatic sarcoma remains as low as < 35%. Currently, there are no biomarkers available to assess disease status in patients with sarcomas and as such, disease surveillance remains only reliant on serial imaging which increases the risk of secondary malignancies and heightens patient anxiety. Methods: Here, we enumerated the cell surface vimentin (CSV+) CTCs in the blood of 92 sarcoma pediatric and adolescent and young adult (AYA) patients. Results: We discovered the CSV+CTC could serve as a possible biomarker of disease with sensitivity of 85.3% and specificity of 75%. Significantly, patients who were deemed to be CSV+ CTC positive were found to have a worse overall survival compared to those who were CSV+ CTC negative (773 days vs 2622 days). Addition of readily available genetic analyses improved the sensitivity in both diagnostic and prognostic tests. Conclusions: Our findings indicate that CSV+ CTCs have both diagnostic and prognostic value and can possibly serve as a measure of disease burden.

Updated outcomes for patients with completely resected pulmonary recurrent osteosarcoma: A report from the Children’s Oncology Group

10023

Background: Amongst patients with recurrent osteosarcoma (OS), those with resectable pulmonary-only relapse appear to have the best outcomes. Prior analysis of patients with completely resected recurrent OS enrolled on a Children’s Oncology Group (COG) Phase 2 trial AOST0221, which studied the efficacy of inhaled GMCSF, showed a 12- month disease control rate (DCR12) of only 20% (95% CI, 10 – 34%). DCR12 based on this analysis was used as the historical benchmark for efficacy analysis in two recently completed trials, AOST1321 (completely resected cohort only) and AOST1421. We analyzed the stability of the DCR12 benchmark using data from these contemporary studies. Methods: Patients were eligible for AOST1321 if they had undergone resection of all sites of recurrent or refractory OS within 30 days of enrollment and for AOST1421 if they had lung only recurrent OS completely resected within 4 weeks of enrollment. AOST1321 evaluated denosumab while AOST1421 evaluated dinutuximab. Patients with refractory disease, extrapulmonary recurrence or without histological confirmation of relapse were excluded from this analysis. DCR12 was defined as having at least stable disease 12 months after the start of protocol therapy. We report the DCR12 observed on AOST1321, AOST1421 and AOST0221. Results: One hundred and twenty-eight evaluable patients with completely resected recurrent OS were enrolled on AOST0221 (N=49), AOST1321 (N=38) and AOST1421 (N=41). One hundred and one patients were included in this analysis (AOST0221: 37, AOST 1321: 25, AOST1421:39). DCR12 was 14 % (95% CI, 5% – 26%) for AOST0221, 24% (95%CI 10-42%) for AOST1321, and 31% (95% CI, 17% - 45%) for AOST1421. Risk for disease progression did not differ across the 3 included studies. DCR12 for all three studies combined was 23% (95% CI, 15% – 31%). Conclusions: Prognosis for recurrent OS remains dismal, even for those with resectable pulmonary-only disease. Although not statistically significant, dinutuximab may have activity in a select group of relapsed OS patients; combination studies using dinutuximab are planned. The previously described benchmark of disease control at 12 months remained relatively consistent in recent studies for those patients who have resectable pulmonary-only relapse. Therefore, DCR12 remains a useful outcome measure in fully resected OS with lung only metastases. Analyses are ongoing to better define the appropriate threshold to define investigational agent activity in this specific patient population.

A phase I trial of aerosol gemcitabine for the treatment of patients with solid tumors and lung metastases

TPS3645

Background: Pre-clinical studies of aerosol gemcitabine (GCB) in mice and dogs with osteosarcoma (OS) lung metastases demonstrated therapeutic efficacy. Aerosol GCB administered once weekly proved to be safe in adults with lung cancer. Direct delivery of GCB to the lungs via inhalation may offer higher drug concentration in the tumor with fewer side effects. We initiated a Phase I study to evaluate the feasibility and safety of aerosol GCB treatment in patients >12 years with solid tumors and lung metastases (2015-0720- NCT03093909). Methods: Eligibility criteria: 1) Diagnosis of solid tumor with lung metastases, 2) willing to comply with protocol therapy, 3) adequate organ function, 4) patient age > 12 and < 50 years, 5) good performance status, 6) resolution of all acute toxic effects of any prior anti-cancer therapy, and 7) no radiotherapy within 2 weeks. Patients who previously received systemic GCB are eligible. Objectives: To determine the maximum tolerated dose (MTD) and toxicities of aerosol GCB, to evaluate for drug spillover into the circulation, and to preliminarily assess the anti-tumor activity. Correlative studies include effect of aerosol GCB on immune cell infiltration in the lung, autophagy, apoptosis, heat shock protein 27, evidence of DNA strand breaks (gH2AX) and expression of human equilibrative nucleoside transporter-1.Aerosol GCB is administered via a breath-induced nebulizer twice a week in 28-day cycles. A maximum of 6 dose levels will be studied; the starting dose is 0.75 mg/kg twice weekly. If no progressive disease or unacceptable treatment-related toxicity, patients may continue for 12 cycles. The study uses the accelerated titration method for the first 2 dose levels then the 3+3 design for the remaining dose levels. After determining the MTD, we will evaluate the defined MTD in an expansion cohort of 14 patients with relapsed OS. Symptoms, pulse oximetry, and pulmonary function are assessed prior to each nebulized dose using remote spirometry that allows raw numbers and flow-volume curves to be uploaded and transmitted via bluetooth to an android tablet provided to patients. Data is transmitted to a web portal and captured in a HIPAA-compliant web-based database (REDCap) that is accessible to the research team. Results: To date, the study enrolled 4 patients and accrual is ongoing at dose level 3. Conclusions: This study will provide information on the feasibility and safety of aerosol GCB. If proven to be feasible and safe, it can potentially offer a novel approach to treat metastatic OS to the lungs while minimizing systemic toxicity. Clinical trial information: NCT03093909 .

Phase II study of antidisialoganglioside antibody, dinutuximab, in combination with GM-CSF in patients with recurrent osteosarcoma (AOST1421): A report from the Children’s Oncology Group

10508

Background: Treatment of patients with recurrent osteosarcoma (OS) is challenging and novel effective therapies are urgently needed. This study evaluated disease control rate (DCR) in patients with recurrent pulmonary OS, when treated with dinutuximab plus cytokine therapy as compared to a historical benchmark. The rationale for dinutuximab was the ubiquitous ( > 95%) GD2 positivity in OS tumors and cell lines. Methods: AOST1421 was a single-arm phase 2 study. Patients with recurrent pulmonary OS in complete surgical remission were eligible. Patients received five cycles of dinutuximab 70mg/m2/cycle with GM-CSF. Two different dinutuximab infusion schedules were used - 35mg/m2/day over 20 hours (2-day) and 17.5mg/m2/day over 10 hours (4-day) schedule. Primary end point was DCR, defined as proportion of patients event-free at 12 months from enrollment. Events were progressive disease or death within 12 months attributed to treatment or progression. The historical benchmark was AOST0221 with a 12-month DCR of 20% (95% CI 10-34%). Success was defined as ≥16/ 39 patients ( > 40%) event-free at 12 months from enrollment. Secondary objectives included toxicity evaluation and dinutuximab pharmacokinetics (PK). Results: Forty-one patients were enrolled from Nov 2015 - Jan 2018. Thirty nine were eligible and evaluable (age 7-26 yr; median 15 yr). Data current to December 31, 2019 was used for analysis. Accrual rate was higher than expected (22.1 vs. 19.2 patients/ yr.) despite a concurrently open competing study. One of 136 administered therapy cycles met criteria for unacceptable toxicity when one patient receiving the 2-day schedule died after cycle 2 due to an unknown cause, attributed as probably related to protocol therapy. The protocol was revised to allow only the 4-day schedule. Other ≥ Grade 3 toxicities occurring in > 10 % participants were expected dinutuximab toxicities such as pain, diarrhea, hypoxia and hypotension. Dinutuximab did not demonstrate sufficient evidence of efficacy as 27/ 39 patients experienced an event for a DCR of 30.7% (95% CI 17- 47%). PK studies are pending and will be reported. Conclusions: Dinutuximab toxicity in adolescent and young adult OS patients was similar to younger patients. While GD2 remains a relevant target in OS, combination of dinutuximab with GM-CSF did not meet the targeted successful DCR in patients with completely resected tumor. Other strategies for targeting GD2 or dinituximab combination therapy may still be warranted. Clinical trial information: NCT02484443.

Maximum tumor dimension and tumor volume as prognostic factors in patients with newly diagnosed localized Ewing sarcoma (ES)- a report from the Children’s Oncology Group (COG)

11529

Background: Maximum tumor dimension > 8 cm. and large tumor volume have been reported to be adverse prognostic factors in patients with ES but have not been prospectively evaluated in the context of a phase 3 clinical trial with interval compressed chemotherapy. Methods: COG AEWS1031 (NCT01231906) was a randomized phase 3 clinical trial comparing interval compressed chemotherapy regimens in patients with newly diagnosed localized ES of bone and soft tissue. A correlative objective of AEWS1031 was to evaluate tumor size and volume as prognostic factors. Institution-reported dimensions of the primary tumor from baseline imaging were prospectively collected. For inclusion in this analysis, patients had to have at least 1 tumor dimension reported for tumor size analyses and dimensions in 3 axes for tumor volume analyses. Maximum dimension was dichotomized as less than vs. > / = 8cm. Tumor volume was dichotomized as less than vs. > / = 200 mL. Event-free (EFS) and overall survival (OS) from enrollment were calculated using Kaplan-Meier methods and compared between groups using a two-sided log-rank test. Hazard ratios (HR) and confidence intervals (CI) were calculated using the Cox model. Results: The 5-year EFS and OS of the 629 eligible patients was 78% (95% CI: 75-81%) and 87% (95% CI: 84-90%) respectively and there was no significant difference in both EFS and OS between the randomized interval compressed chemotherapy arms of AEWS1031. 590 of 629 (94%) patients were evaluable for maximum tumor dimension and 307 (52%) had tumors > / = 8 cm. Patients with tumors > / = 8 cm were at significantly increased risk for EFS events (p = 0.016) with estimated 5-year EFS of 73.7% (95% CI: 68.1 vs.78.4%) vs. 82.9% (95% CI 77.7-87.1%) for patients with tumors < 8 cm [HR: 1.53 (1.08-2.17)]. For tumor volume, 586 of 629 patients (93%) were evaluable and 180 (31%) had tumors > / = 200 mL. Patients with tumor volume > / = 200 mL were at significantly increased risk for EFS events (p = 0.003) with estimated 5-year EFS of 70% (95% CI: 62.3-76.4%) vs. 81.6% (95% CI: 77.2-85.2%) for patients with tumors < 200 mL [HR: 1.69 (1.2-2.39)]. Conclusions: Maximum tumor dimension and tumor volume as defined are both prognostic in patients with newly diagnosed localized ES treated with interval compressed chemotherapy. Clinical trial information: NCT01231906 .

Evaluation of the TPO-receptor agonist Eltrombopag in the Pediatric Preclinical Testing Consortium osteosarcoma in vivo models

e22502

Background: Eltrombopag (EP) is a small molecule, thrombopoietin receptor (TPO-R) agonist indicated for the treatment of patients with chronic immune thrombocytopenia and severe aplastic anemia. EP is a polyvalent cation chelator and inhibits leukemia cell proliferation via depletion of intracellular iron. Recent studies show EP inhibits the proliferation of osteosarcoma cells lines via depletion of polyvalent cations. The in vivo effects of EP were studied in osteosarcoma patient derived xenograft models by the Pediatric Preclinical Testing Consortium (PPTC). Methods: The in vivo anticancer effects of EP were assessed in a panel of six osteosarcoma PPTC PDX models with limited MPL mRNA expression (OS2, OS9, OS31, OS33, OS36, OS60). EP was administered at an oral dose of 5 mg/kg/day given for 5 days each week with planned treatment period of 4 weeks. High dose EP (50mg/kg/day) was also tested in 2 PDX models (OS2, OS9) on the same schedule. A control cohort that received vehicle was included for each PDX model. Tumor volumes were measured and responses defined utilizing the PPTC statistical analyses. Results: EP at 5 mg/kg failed to inhibit tumor growth or induce significant differences in event-free survival (EFS) in any of the 6 osteosarcoma PDX models. At the higher dose of 50 mg/kg a significant prolongation in time to event in the EP-treated group was observed, but the effect was small with the ratio of the median time to event for treated versus control animals (EFS T/C) being only 1.2 in the 2 OS PDX models tested. No objective responses were observed with EP at either dose, with all models demonstrating progressive disease. Conclusions: EP did not exhibit significant antitumor activity against the PPTC osteosarcoma PDX panel. However, EP also did not enhance tumor growth. EP’s lack of anti-tumor activity against the OS PDX models suggests leukemia and osteosarcoma cells likely have different dependencies on intracellular polyvalent cations. Given EPs effect on stimulating platelet production, and the demonstration that EP does not stimulate in vivo growth of osteosarcoma, EP may be considered in patients with osteosarcoma as a supportive care agent in supporting platelet recovery.

Evaluation of the multi-kinase inhibitor regorafenib in the Pediatric Preclinical Testing Consortium osteosarcoma, rhabdomyosarcoma, and Ewing sarcoma in vivo models

10038

Background: Regorafenib is a multi-kinase inhibitor, developed by adding a fluorine atom to the phenyl ring of sorafenib. Regorafenib inhibits multiple kinases including BRAF, FGFR1, KIT, PDGFRB, RAF, RET, and VEGFR1-3, many at a higher potency than sorafenib. Prior studies within the Pediatric Preclinical Testing Consortium (PPTC) demonstrated sorafenib exhibited intermediate activity for tumor growth inhibition in more than 50% of the sarcoma models tested at a dose of 60mg/kg by oral gavage daily (5 days/wk for 6 consecutive weeks). The in vivo effects of regorafenib were studied in the PPTC osteosarcoma (OS), rhabdomyosarcoma (Rh) and Ewing (EW) sarcoma xenograft models. Methods: The in vivo anticancer effects of regorafenib were assessed in a panel of 6 osteosarcoma models (OS2, OS9, OS31, OS33, OS36, OS60), two rhabdomyosarcoma models (Rh30, Rh41), and one Ewing sarcoma model (EW5). Regorafenib was administered by oral gavage at a dose of 30 mg/kg/day given daily for 21 consecutive days. Time to event and tumor volume responses were defined and analyzed utilizing standard PPTC statistical methods. Results: Regorafenib induced significant improvements in event-free survival (EFS) compared to control in 100% (9/9) of sarcoma models tested. Most models showed pronounced slowing of tumor growth compared to control during the 21 days of regorafenib treatment, with tumor growth generally approximating control rates soon after completion of regorafenib treatment. Three out of 8 sarcoma models demonstrated EFS T/C values > 2 (1/6 OS, 2/2 Rh, 0/1 EW). Minimum relative tumor volumes ranged from 0.74 to 1.60, with no models meeting criteria for objective response. Conclusions: Regorafenib induced modest inhibition of tumor growth in the PPTC sarcoma models evaluated. The overall pattern of response to the multi-kinase inhibitor regorafenib against the PPTC sarcoma models appears similar to that of the kinase inhibitor sorafenib, with pronounced slowing of tumor growth in some models that is limited to the period of agent administration being the primary treatment effect.

A DNA methylation-based classifier for accurate molecular diagnosis of bone sarcomas

11034

Background: Bone sarcomas present a unique diagnostic challenge because of the considerable morphologic overlap between different entities. The choice of optimal treatment, however, is dependent upon accurate diagnosis. Genome-wide DNA methylation profiling has emerged as a new approach to aid in the diagnosis of brain tumors, with diagnostic accuracy exceeding standard histopathology. In this work we developed and validated a methylation based classifier to differentiate between osteosarcoma, Ewing’s sarcoma, and synovial sarcoma. Methods: DNA methylation status of 482,421 CpG sites in 15 osteosarcoma, 10 Ewing’s sarcoma, and 11 synovial sarcoma samples were measured using the Illumina HumanMethylation450 array. From this training set of 36 samples we developed a random forest classifier using the 400 most differentially methylated CpG sites (FDR q value < 0.001). This classifier was then validated on 10 synovial sarcoma samples from TCGA, 86 osteosarcoma samples from TARGET-OS, and 15 Ewing’s sarcoma from a recently published series (Huertas-Martinez et al., Cancer Letters 2016). Results: Methylation profiling revealed three distinct molecular clusters, each enriched with a single sarcoma subtype. Within the validation cohorts, all samples from TCGA were correctly classified as synovial sarcoma (10/10, sensitivity and specificity 100%). All but one sample from TARGET-OS were classified as osteosarcoma (85/86, sensitivity 98%, specificity 100%) and all but one sample from the Ewing’s sarcoma series was classified as Ewing’s sarcoma (14/15, sensitivity 93%, specificity 100%). The single misclassified osteosarcoma sample was classified as Ewing’s sarcoma, and was later determined to be a misdiagnosed Ewing’s sarcoma based on RNA-Seq demonstrating high EWRS1 and ETV1 expression. An additional clinical sample that was misdiagnosed as a synovial sarcoma by initial histolopathology, was accurately recognized as osteosarcoma by the methylation classifier. Conclusions: Osteosarcoma, Ewing’s sarcoma and synovial sarcoma have distinct epigenetic profiles. Our validated methylation-based classifier can be used to provide an accurate diagnosis when histological and standard techniques are inconclusive.

Complete dexrazoxane cardioprotection for cardiac function but incomplete female cardioprotection for cardiac structure in doxorubicin-treated osteosarcoma survivors: Hearts too small for the body

10519

Background: Dexrazoxane is protective for lower-dose doxorubicin ( < 300 mg/m2) cardiotoxicity in childhood cancer, but the effect of dexrazoxane (DXRZ) administered with higher-dose (HD) doxorubicin (DOXO) is unknown. Methods: We evaluated patients from Children’s Oncology Group trials for localized (P9754) and metastatic (AOST0121) osteosarcoma (OS) who received HD DOXO (375-600 mg/m2) preceded by DXRZ (10:1 ratio), methotrexate, and cisplatin; some also received ifosfamide alone or ifosfamide/etoposide ± trastuzumab. Cardiotoxicity was identified by echocardiography and by serum N-terminal pro-brain natriuretic peptide (NT-proBNP) concentrations. Results: 81 DXRZ -treated OS patients ( age at enrollment = 13.7 years; range 3.8 - 23.7 years) had normal left ventricular (LV) systolic function as measured by LV fractional shortening and no heart failure. Female sex and longer follow-up since DOXO were associated with a significantly smaller LV dimension z-score normalized to BSA (μ = -1.20, 95%CI [-1.70, -0.70]). Similarly, in the one-third of patients treated > 81 days after minimal expected treatment (groups equally partitioned by time), significantly thinner LV posterior wall thickness for BSA (μ = -0.57, [-1.05, -0.09]) was found. Interventricular septal wall thickness (μ = -0.84, [-1.2, -0.48]) and LV mass (μ = -0.73, [-1.06, -0.40]) were significantly smaller for BSA than normal for both sexes. For females, these became significantly more abnormal with increasing length of follow-up. Females also showed progressive increases in NT-proBNP. Conclusions: DXRZ is cardioprotective for HD DOXO in terms of LV function and heart failure. Females had progressive abnormalities of LV structure, leading to smaller hearts for body size. This was associated with increasing cardiac stress, as measured by NT-proBNP. DXRZ protection was incomplete for HD DOXO effects on LV structure, resulting in higher LV stress and risk for late LV dysfunction. DXRZ should continue to be used in this population, including for females who exhibit more cardiotoxicity than males at specific cumulative DOXO doses.

MAP plus maintenance pegylated interferon α-2b (MAPIfn) versus MAP alone in patients with resectable high-grade osteosarcoma and good histologic response to preoperative MAP: First results of the EURAMOS-1 “good response” randomization

LBA10504

Background: EURAMOS-1 (NCT00134030) is an international randomized controlled trial investigating treatment optimization in osteosarcoma on the basis of histologic response to preoperative chemotherapy (CT). In patients (pts) with “good response” (GR; <10% viable tumor at surgery) to induction CT, efficacy of maintenance therapy with pegylated interferon α-2b (Ifn) was investigated. Methods: Pts ≤40yrs with resectable localized or primary metastatic high-grade extremity or axial osteosarcoma were eligible for registration at diagnosis. Eligibility for randomization included: [a] receipt of two cycles pre-op methotrexate, doxorubicin, cisplatin (MAP); [b] complete macroscopic resection of primary tumor; and [c] no disease progression. GR pts were randomized (1:1) after surgery to four cycles MAP +/-Ifn. Ifn (0.5-1.0 μg/kg/wk) was given after CT until 2yr post registration. Primary endpoint: event free survival (EFS); secondary: overall survival, toxicity. The trial design sought improved 3yr EFS from 70% to 80% (80% power, 2-sided alpha 5%). The final analysis was planned after 147 EFS events. Results: 2,260 pts were registered (Apr 05 to Jun 11) from 17 countries (326 sites) making it the largest trial in this rare cancer. GR was reported in 1,034 pts and 715 consented to randomization (358 MAP, 357 MAPIfn) with baseline characteristics balanced by arm: median age 14yr, 59% male, 79% localized disease. 271/357 (76%) pts started Ifn; main reason for not starting was treatment refusal, 66/86. By Feb 13 234/271 had stopped Ifn. Median duration of Ifn if started was 455 days; grade 3+ toxicity during Ifn reported by 81pts (30%). Median follow-up is 3.1yr with 174 EFS events reported: 93 MAP, 81 MAPIfn. Hazard ratio for EFS from adjusted Cox model was 0.82 (95% CI 0.61-1.11; p=0.201) in favor of MAPIfn. Conclusions: With current follow-up, EFS for MAPIfn is not statistically superior to MAP alone in pts with high-grade osteosarcoma and good response to pre-op MAP. One-quarter of pts did not start Ifn, which may have influenced this finding. Further follow-up for events and survival continues. Clinical trial information: NCT00134030.

Discordance of oncologic surgical classifications in COG studies

10528

Background: “Wide resection” (a cuff of normal tissue) versus “radical resection” (the entire compartment) indicate very distinct oncologic surgical procedures and hence potential margin status. Distinguishing between these two oncologic classifications is important for understanding oncologic outcomes. Methods: We examined the available data for COG AOST0331: A Randomized Trial of the European and American Osteosarcoma Study Group to Optimize Treatment Strategies for Resectable Osteosarcoma Based on Histological Response to Pre-Operative Chemotherapy. We reviewed the surgical and pathology reports of patients considered to have received a wide or radical resection according to the investigator at the patient’s institution. Results: In 956 patients, the overall discordance rate was 43%. Of those patients reported to have had a wide excision by the reporting institution, only 5% of patients were reclassified to radical resection. However, of those patients reported to have undergone radical resection, 75% were reclassified to wide. In the absence of this re-review of the data, 56% of the patients would have been reported to have had a radical resection when in fact only 17% met the criteria for true radical resection. The greater number of patient reclassified to wide from radical is likely indicative of the influence and limits of the CPT coding system currently used for billing by surgeons; only CPT codes for radical resection currently exist. Conclusions: These discrepancies complicate subsequent data analysis, particularly assessing predictors of local recurrence. While this reporting problem is not unique to COG studies, we aim to use this example to raise awareness among our colleagues about the importance of accurate data reporting. Providing better training to those individuals responsible for submitting data may decrease the incidence of erroneous data entry and is essential to the successful completion of study objectives.

MAP plus maintenance pegylated interferon α-2b (MAP-IFN) versus MAP alone in patients (pts) with resectable high-grade osteosarcoma and good histologic response to preoperative MAP: First results of the EURAMOS-1 good response randomization

LBA10504

The full, final text of this abstract will be available at abstract.asco.org at 7:30 AM (EDT) on Monday, June, 3, 2013, and in the Annual Meeting Proceedings online supplement to the June 20, 2013, issue of Journal of Clinical Oncology. Onsite at the Meeting, this abstract will be printed in the Monday edition of ASCO Daily News.

The Children’s Oncology Group and QuadW Foundation osteosarcoma banking experience

10053

Background: Survival rates of osteosarcoma patients have remained stagnant over the last twenty years. Predictive biomarkers of response to treatment and likelihood of recurrence are yet to be developed. In order to develop new therapeutics and better risk-stratified regimens, a greater understanding of osteosarcoma biology is needed. The Children’s Oncology Group (COG) addressed this need by banking more than 10,000 tumor and tissue samples from over 1500 patients with osteosarcoma, but many samples lacked clinically relevant data. Methods: 1105 eligible patients enrolled on the COG osteosarcoma biology study P9851. Of those, 510 patients were not enrolled on a concurrent therapeutic trial, which limited the clinical annotation of those samples. 589 patients have enrolled on the successor study AOST06B1. The lack of clinical annotation of the P9851 specimens limited their value, and the lack of statistical support slowed the analysis of several biology studies. The QuadW Foundation, CureSearch, and the COG formed the Childhood Sarcoma Biostatistics and Annotation Office (CSBAO) in 2008 to link clinically annotated patient data to the samples and to provide statistical support. Results: In 2008, only 5.3% of samples from the 510 P9851 patients not enrolled on a therapeutic study had full clinical annotation. The efforts of the CSBAO have linked clinical annotation to 90.8% of those specimens and provided statistical analyses. As a result, 18 biology studies in osteosarcoma are completed, with 11 published in peer-reviewed journals. Samples were provided to the TARGET program (for gene expression arrays, copy number analysis, and sequencing data); the resulting data will be available to the scientific community for in silico studies. Conclusions: The efforts of the CSBAO have led to a substantial increase in the value of a large osteosarcoma biospecimen repository by clinically annotating the specimens and providing statistical resources for analyses of planned and completed studies. These samples with annotated patient information are available by request to the research community for basic and translational science projects to improve the biological understanding of, and the treatment of patients with osteosarcoma. Clinical trial information: NCT00899275.

The effects of dexrazoxane on cardiac status and second malignant neoplasms (SMN) in doxorubicin-treated patients with osteosarcoma (OS)

9503

Background: Anthracyclines are highly efficacious in OS but associated with risk of cardiotoxicity. We conducted two studies using dexrazoxane as a cardioprotectant: i) AOST0121, a phase II study for metastatic OS and ii) P9754, a series of pilot studies including doxorubicin dose intensification in patients with localized OS. Methods: Patients on AOST0121 received methotrexate, doxorubicin (375 mg/m2), cisplatin, ifosfamide and etoposide (MAPIE). Those with HER2-positive tumors also received trastuzumab (4 mg/kg loading dose, 2 mg/kg weekly). On P9754, patients received MAP in Pilot 1, MAPI or MAPIE in Pilots 2 and 3. Total doxorubicin was 450 mg/m2 or 600 mg/m2 with each dose preceded by dexrazoxane (10:1 ratio). Etoposide was never given simultaneously with dexrazoxane. Measurements of left ventricular systolic function by echocardiography, serum troponin-T (cTnT, a myocardial injury marker) and N-terminal pro-brain natriuretic peptide (NT-proBNP, a cardiomyopathy marker) were obtained at baseline and repeated during therapy. Secondary malignancies were reported to the NCI. Results: None of the 47 patients (17 receiving trastuzumab) evaluated for cardiac toxicity on AOST0121 had significant changes in left ventricular fractional shortening, measurable cTnT, elevation of NT-proBNP or clinical evidence of cardiotoxicity (CTCv2.0). In P9754, 242 patients were evaluated for cardiotoxicity. Only 1 patient had CTCv2.0 grade 3 toxicity and 1 patient had a measurable cTnT. Three of 96 patients treated on AOST0121 and two of 272 patients on P9754 developed AML. Combining both studies: 5/398 or 1.4%. Conclusions: This large group of OS patients demonstrates that dexrazoxane is an effective cardioprotectant for doxorubicin alone (375-600 mg/m2), and in combination with trastuzumab. Dexrazoxane did not lead to an increase in secondary malignancies in OS patients treated with these regimens as compared to historical rates of 1-2%. Our results do not support the findings of a recent European Medicines Agency safety review that concluded dexrazoxane was unsafe for use in children and adolescents due to risk of SMN.

PHC3, a component of the hPRC-H complex, associates with E2F6 during G0 and is lost in osteosarcoma tumors

Polyhomeotic-like 3 (PHC3) is a ubiquitously expressed member of the polycomb gene family and part of the human polycomb complex hPRC-H. We found that in normal cells PHC3 associated with both hPRC-H complex components and with the transcription factor E2F6. In differentiating and confluent cells, PHC3 and E2F6 showed nuclear colocalization in a punctate pattern that resembled the binding of polycomb bodies to heterochromatin. This punctate pattern was not seen in proliferating cells suggesting that PHC3 may be part of an E2F6-polycomb complex that has been shown to occupy and silence target promoters in G(0). Previous loss of heterozygosity (LoH) analyses had shown that the region containing PHC3 underwent frequent LoH in primary human osteosarcoma tumors. When we examined normal bone and human osteosarcoma tumors, we found loss of PHC3 expression in 36 of 56 osteosarcoma tumors. Sequence analysis revealed that PHC3 was mutated in nine of 15 primary osteosarcoma tumors. These findings suggest that loss of PHC3 may favor tumorigenesis by potentially disrupting the ability of cells to remain in G(0).

High-dose melphalan, etoposide, total-body irradiation, and autologous stem-cell reconstitution as consolidation therapy for high-risk Ewing's sarcoma does not improve prognosis

PURPOSE

To determine whether consolidation therapy with high-dose melphalan, etoposide, and total-body irradiation (TBI) with autologous stem-cell support would improve the prognosis for patients with newly diagnosed metastatic Ewing's sarcoma (ES).

PATIENTS AND METHODS

Thirty-two eligible patients with newly diagnosed ES metastatic to bone and/or bone marrow were enrolled onto this study. Treatment was initially comprised of five cycles of induction chemotherapy (cyclophosphamide, doxorubicin, and vincristine alternating with ifosfamide and etoposide) and local control. Peripheral-blood stem-cell collection was performed after the second cycle of chemotherapy, with delay if the bone marrow was persistently involved. If patients had a good response to initial therapy, they proceeded to consolidation therapy with melphalan, etoposide, TBI, and stem-cell support.

RESULTS

Of the 32 eligible patients, 23 proceeded to high-dose therapy consolidation. Of the nine patients who did not proceed to consolidation, four were secondary to progressive disease and two were secondary to toxicity. Three patients died from toxicity during the high-dose phase of the therapy. The majority of the patients who underwent high-dose consolidation therapy experienced relapse and died with progressive disease. Two-year event-free survival (EFS) for all eligible patients is 20%. The 2-year post-stem-cell reconstitution EFS for the subset of 23 patients who received consolidation therapy is 24%. Analysis of peripheral-blood stem-cell collections by molecular techniques for minimal residual disease showed contamination of at least some samples by tumor cells in all three patients with available data.

CONCLUSION

Consolidation with high-dose melphalan, etoposide, TBI, and autologous stem-cell support failed to improve the probability of EFS in this cohort of patients with newly diagnosed metastatic ES.