Search

PURPOSE Adaptive clinical trials use algorithms to predict, during the study, patient outcomes and final study results. These predictions trigger interim decisions, such as early discontinuation of the trial, and can change the course of the study. Poor selection of the Prediction Analyses and Interim Decisions (PAID) plan in an adaptive clinical trial can have negative consequences, including the risk of exposing patients to ineffective or toxic treatments. METHODS We present an approach that leverages data sets from completed trials to evaluate and compare candidate PAIDs using interpretable validation metrics. The goal is to determine whether and how to incorporate predictions into major interim decisions in a clinical trial. Candidate PAIDs can differ in several aspects, such as the prediction models used, timing of interim analyses, and potential use of external data sets. To illustrate our approach, we considered a randomized clinical trial in glioblastoma. The study design includes interim futility analyses on the basis of the predictive probability that the final analysis, at the completion of the study, will provide significant evidence of treatment effects. We examined various PAIDs with different levels of complexity to investigate if the use of biomarkers, external data, or novel algorithms improved interim decisions in the glioblastoma clinical trial. RESULTS Validation analyses on the basis of completed trials and electronic health records support the selection of algorithms, predictive models, and other aspects of PAIDs for use in adaptive clinical trials. By contrast, PAID evaluations on the basis of arbitrarily defined ad hoc simulation scenarios, which are not tailored to previous clinical data and experience, tend to overvalue complex prediction procedures and produce poor estimates of trial operating characteristics such as power and the number of enrolled patients. CONCLUSION Validation analyses on the basis of completed trials and real world data support the selection of predictive models, interim analysis rules, and other aspects of PAIDs in future clinical trials.

PURPOSE The use of growth factors adds considerable expense and some toxicity to adjuvant breast cancer chemotherapy. We tested the feasibility and safety of omitting routine peg-filgrastim use during the paclitaxel portion of the dose-dense doxorubicin-cyclophosphamide–paclitaxel regimen. PATIENTS AND METHODS This was a prospective, single-arm study in which patients 18 to 65 years of age who completed 4 cycles of dose-dense doxorubicin-cyclophosphamide for stage I-III breast cancer received paclitaxel 175 mg/m2 every 2 weeks. Peg-filgrastim was administered after paclitaxel only if patients had had febrile neutropenia in a prior cycle or at investigator discretion if patients had infections or treatment delays of > 1 week. Once a patient received peg-filgrastim, it was administered in all future cycles. The primary end point was the rate of paclitaxel completion within 7 weeks from cycle 1 day 1 to cycle 4 day 1. If ≥ 100 out of 125 patients completed 4 cycles of paclitaxel without dose delay, the regimen would be considered feasible. RESULTS The enrollment goal of 125 patients was met. Median age was 46 years (range, 21-65 years), and 112 patients (90% [95% CI, 83% to 94%]) completed dose-dense paclitaxel within 7 weeks. Omission of peg-filgrastim was not causally related to noncompletion of paclitaxel in any patients. The most common reasons for dose reduction or delays were nonhematologic. One patient experienced febrile neutropenia but was able to complete paclitaxel on time. Eight patients (6.4%) received peg-filgrastim during the trial. Overall, peg-filgrastim was administered in only 4.3% of paclitaxel cycles. CONCLUSION Omission of routine peg-filgrastim during dose-dense paclitaxel according to a prespecified algorithm seems to be safe and feasible and was associated with a 95.7% reduction in the use of peg-filgrastim relative to the current standard of care.

PURPOSE Smoldering multiple myeloma (SMM) is a precursor condition of multiple myeloma (MM) with a 10% annual risk of progression. Various prognostic models exist for risk stratification; however, those are based on solely clinical metrics. The discovery of genomic alterations that underlie disease progression to MM could improve current risk models. METHODS We used next-generation sequencing to study 214 patients with SMM. We performed whole-exome sequencing on 166 tumors, including 5 with serial samples, and deep targeted sequencing on 48 tumors. RESULTS We observed that most of the genetic alterations necessary for progression have already been acquired by the diagnosis of SMM. Particularly, we found that alterations of the mitogen-activated protein kinase pathway ( KRAS and NRAS single nucleotide variants [SNVs]), the DNA repair pathway (deletion 17p, TP53, and ATM SNVs), and MYC (translocations or copy number variations) were all independent risk factors of progression after accounting for clinical risk staging. We validated these findings in an external SMM cohort by showing that patients who have any of these three features have a higher risk of progressing to MM. Moreover, APOBEC associated mutations were enriched in patients who progressed and were associated with a shorter time to progression in our cohort. CONCLUSION SMM is a genetically mature entity whereby most driver genetic alterations have already occurred, which suggests the existence of a right-skewed model of genetic evolution from monoclonal gammopathy of undetermined significance to MM. We identified and externally validated genomic predictors of progression that could distinguish patients at high risk of progression to MM and, thus, improve on the precision of current clinical models.

PURPOSE The yield of comprehensive genomic profiling in recruiting patients to molecular-based trials designed for small subgroups has not been fully evaluated. We evaluated the likelihood of enrollment in a clinical trial that required the identification of a specific genomic change based on our institute-wide genomic tumor profiling. PATIENTS AND METHODS Using genomic profiling from archived tissue samples derived from patients with metastatic breast cancer treated between 2011 and 2017, we assessed the impact of systematic genomic characterization on enrollment in an ongoing phase II trial (ClinicalTrials.gov identifier: NCT01670877 ). Our primary aim was to describe the proportion of patients with a qualifying ERBB2 mutation identified by our institutional genomic panel (OncoMap or OncoPanel) who enrolled in the trial. Secondary objectives included median time from testing result to trial registration, description of the spectrum of ERBB2 mutations, and survival. Associations were calculated using Fisher’s exact test. RESULTS We identified a total of 1,045 patients with metastatic breast cancer without ERBB2 amplification who had available genomic testing results. Of these, 42 patients were found to have ERBB2 mutation and 19 patients (1.8%) were eligible for the trial on the basis of the presence of an activating mutation, 18 of which were identified by OncoPanel testing. Fifty-eight percent of potentially eligible patients were approached, and 33.3% of eligible patients enrolled in the trial guided exclusively by OncoPanel testing. CONCLUSION More than one half of eligible patients were approached for trial participation and, significantly, one third of those were enrolled in NCT01670877. Our data illustrate the ability to enroll patients in trials of rare subsets in routine clinical practice and highlight the need for these broadly based approaches to effectively support the success of these studies.

2051 Background: Randomized clinical trials use stringent eligibility criteria to select patients which can raise concerns about generalizability of study results. Recent interest in external controls and new trial designs has increasingly focused on the possible use of real-world data sources. We examined potential differences of survival outcomes in clinical trial and real-world datasets in newly diagnosed glioblastoma. Methods: Patient-level data from 4 independent datasets were analyzed. Non-trial data were derived from an institutional dataset of 453 patients with newly diagnosed glioblastoma treated outside of a clinical trial with standard radiation therapy with concurrent and adjuvant temozolomide at a large academic center. Trial patients included patients on the control arm of several multi-institutional trials (NCT00689221, n = 273; NCT00813943, n = 89; NCT00943826; n = 459) and patients treated on trial at our institution (n = 210). Non-trial patients were compared with each of the 4 clinical trial datasets in pairwise comparisons with Cox regression with adjustments for age, sex, extent of resection, KPS and MGMT status. Results: Patient-level data from 1,484 patients were analyzed. Non-trial patients were older compared to patients in the multi-institutional trials (mean 58.6 vs. 56.1 years (NCT00943826), 53.9 years (NCT00813943), 55.7 years (NCT00689221); p < = 0.001). The non-trial cohort had fewer women (43.5% vs. 35.9%, p = 0.02), greater proportion of lower KPS patients (47% KPS 0.05). Conclusions: Glioblastoma patients treated on multi-institutional clinical trials that were included in our analysis did not have statistically significant differences in survival compared to patients treated outside of a clinical trial at a large academic center after adjustment for relevant variables. Our findings support the possible use of real-world data in the development of external control arms for future trials in newly diagnosed glioblastoma. [Table: see text]

512 Background: Sacituzumab govitecan (SG), a novel antibody-drug conjugate in which the topoisomerase 1 inhibitor SN-38 (active metabolite of irinotecan) is linked to a humanized monoclonal antibody targeting the tumor antigen Trop2, is currently approved for treatment of patients (pts) with pre-treated metastatic triple negative breast cancer (TNBC). We conducted a phase 2 study evaluating neoadjuvant (NA) SG as upfront therapy for pts with localized TNBC (NCT04230109). The primary objective was to assess pathological complete response (pCR) rate in breast and lymph nodes (ypT0/isN0) with SG. Secondary objectives included assessment of radiological response rate, evaluation of the safety and tolerability (CTCAE v5.0) and event-free survival (EFS). Methods: Patients with localized TNBC (tumor size ≥1cm, or any size if node positive) with no prior treatment were eligible. SG was administered IV on Days 1, 8 of each 21-day cycle at a starting dose of 10 mg/kg for 4 cycles. After 4 cycles, patients with biopsy-proven residual disease, considered as no pCR for primary endpoint, had the option to receive additional NA therapy at the discretion of the treating physician. Radiologic response (US or MRI) was defined by RECIST version 1.1 using a composite response of CR & PR. Standard descriptive statistics were utilized, including 95% binomial confidence intervals for all rates estimated. Results: From 7/14/20 – 8/31/21, 50 pts were enrolled (median age = 48.5; 11 stage I disease, 24 stage II, 11 stage III, 4 unknown; 62% node negative). The majority (98%; n = 49) of pts completed 4 cycles of SG. Overall, the radiological response rate with SG alone was 62% (n = 31, 95% CI 48%, 77%). 26 pts proceeded directly to surgery after SG. Overall, the pCR rate with SG alone was 30% (n = 15/50, 95% CI 18%, 45%). The other 11 pts had RCB-1 (n = 3), RCB-2 (n = 5), and RCB-3 (n = 3) disease, respectively. Of the 24 pts who received additional NA therapy, 6 had a pCR (3 received anthracycline-based regimen, 2 carboplatin/taxane, and 1 docetaxel/cyclophosphamide). Among pts with a germline BRCA mutation (n = 8), 7 proceeded directly to surgery after SG and 6 had a pCR (86%, 95% CI 42%, 99%). The most common AEs with SG were nausea (82%, n = 41), fatigue (78%, n = 39), alopecia (76%, n = 38), neutropenia (58%, n = 29), anemia (36%, n = 18), and rash (48%, n = 24). 6% of pts required dose-reduction. No pts discontinued SG therapy due to disease progression or AEs; 1 discontinued due to minimal response per investigator preference. At the time of data cut-off (1/18/22), no pts experienced disease recurrence. Updated biomarker and EFS results will be presented at the meeting. Conclusions: In the first neoadjuvant trial in TNBC with an ADC, SG demonstrated single agent efficacy in localized TNBC. Further research on optimal duration of SG as well as NA combination strategies, including immunotherapy, are needed. Clinical trial information: NCT04230109.

Purpose The majority of randomized oncology trials are two-arm studies that test the efficacy of new therapies against a standard of care, thereby assigning a large proportion of patients to nonexperimental therapies. In contrast, multiarm studies efficiently share a common control arm while evaluating multiple experimental therapies. A major bottleneck for traditional multiarm trials is the requirement that all therapies—often drugs from different companies—have to be available at the same time when the trial starts. We evaluate the potential gains of a platform design—the rolling-arms design—that adds and removes arms on a rolling basis. Methods We define the rolling-arms design with the goal of minimizing the complexity of random assignment and data analyses of a platform trial. We then evaluate its potential advantages in hormone receptor–positive, human epidermal growth factor receptor 2–negative advanced breast cancer. Multiple pharmaceutical companies currently test CDK4/6 inhibitors in combination with letrozole in independent two-arm trials. We conducted a simulation study to quantify the reduction in sample size, number of patients treated with the standard of care, and the average time to treatment discovery if these therapies had been tested in a rolling-arms trial. Results A rolling-arms platform design with two to five experimental treatments can reduce the overall sample size requirement by up to 30% compared with standard two-arm studies. It assigns up to 60% fewer patients to the control arm compared with five independent trials that test distinct treatments. Moreover, under realistic scenarios, effective experimental treatments are discovered up to 15 months earlier compared with separate two-arm trials. Conclusion The rolling-arms platform design is applicable to a broad variety of diseases, and under realistic scenarios, it is substantially more efficient than standard two-arm randomized trials.

2014 Background: The cyclin D-CDK4/6-Rb pathway is activated in most glioblastomas. Abemaciclib is a potent CDK4/6 inhibitor with good brain penetration approved for HR+/HER2- breast cancer. In order to efficiently evaluate the potential impact of abemaciclib on overall survival (OS) in newly diagnosed glioblastoma and to simultaneously develop information regarding potential genomic biomarker associations, abemaciclib was included as an arm on the Individualized Screening Trial of Innovative Glioblastoma Therapy (INSIGhT) trial. INSIGhT is a phase II platform trial using response adaptive randomization and deep genomic profiling to more efficiently test experimental agents in MGMT unmethylated glioblastoma and potentially accelerate identification of novel therapies for phase III testing. Initial randomization was equal between abemaciclib, control, and two other experimental arms but subsequent randomization was adapted based on efficacy as determined by progression-free survival (PFS). Ineffective arms were discontinued and new arms added by protocol amendment. We report preliminary results for the abemaciclib arm which has completed accrual. Methods: Patients with newly diagnosed MGMT-unmethylated glioblastoma were randomized to receive either radiotherapy with concomitant and adjuvant temozolomide at standard doses or standard radiochemotherapy followed by adjuvant abemaciclib (150-200 mg orally BID). Treatment continued until progression or development of unacceptable toxicities. The primary endpoint was OS which was assessed using the log-rank test estimated via the Kaplan Meier method using a type I error of 5%. The hazard ratio (HR) was estimated using a cox proportional hazards model. Association between abemaciclib efficacy and cyclin D-CDK4/6-Rb pathway genomic alterations was also investigated. Results: There were 142 patients (69 control; 73 treated with abemaciclib). Abemaciclib was generally well-tolerated with no new toxicity signals identified. PFS was significantly longer (HR 0.67; p = 0.03, logrank test) with abemaciclib (median 6.54 months) compared to the control arm (median 5.88 months). For patients with activation of the CDK4 pathway the PFS HR was 0.64 (p-value = 0.04). However, there was no significant improvement in overall survival (HR 0.9; p-value > 0.05) between abemaciclib (median 15.5) compared to the control arm (median 15.5). Conclusions: Abemaciclib was well-tolerated and prolonged PFS but there is no evidence of an overall survival improvement compared to standard radiochemotherapy. Clinical trial information: NCT02977780.

Purpose Biomarker-based clinical trials provide efficiencies during therapeutic development and form the foundation for precision medicine. These trials must generate information on both experimental therapeutics and putative predictive biomarkers in the context of varying pretrial information. We generated an efficient, flexible design that accommodates various pretrial levels of evidence supporting the predictive capacity of biomarkers while making pretrial design choices explicit. Methods We generated a randomization procedure that explicitly incorporates pretrial estimates of the predictive capacity of biomarkers. To compare the utility of this Bayesian basket (BB) design with that of a balanced randomized, biomarker agnostic (BA) design and a traditional basket (TB) design that includes only biomarker-positive patients, we iteratively simulated hypothetical multiarm clinical trials under various scenarios. Results BB was more efficient than BA while generating more information on the predictive capacity of putative biomarkers than both BA and TB. For simulations of hypothetical multiarm trials of experimental therapies and associated biomarkers of varying incident frequency, BB increased power over BA in cases when the biomarker was predictive and when the experimental therapeutic worked in all patients in a variety of scenarios. BB also generated more information about the predictive capacity of biomarkers than BA and categorically relative to TB, which generates no new biomarker information. Conclusion The BB design offers an efficient way to generate information on both experimental therapeutics and the predictive capacity of putative biomarkers. The design is flexible enough to accommodate varying levels of pretrial biomarker confidence within the same platform structure and makes clinical trial design decisions more explicit.

549 Background: The 21-gene Oncotype DX Recurrence Score (RS) is widely used to guide adjuvant chemotherapy decisions in hormone receptor positive (HR+), HER2-negtive (HER2-), lymph node negative (LN-) breast cancer. It’s adoption in lymph node positive (LN+) disease remains controversial. In 2016, we implemented ‘reflex’ RS testing for patients ≤65 years with HR+/HER2- breast cancer including T1/T2 N1 (grade 1 or 2) tumors. Providers can also order Oncotype DX outside of reflex criteria. We sought to assess RS distribution and factors associated with chemotherapy use in HR+/HER2-/LN+ breast cancer patients at our center. Methods: Patients with non-metastatic HR+/HER2-/LN+ breast cancer who underwent primary surgery at our center were identified from our prospective database. We examined the distribution of low (RS < 18), intermediate (RS 18-30) and high (RS > 30) RS and identified characteristics for those who did not meet reflex criteria. A multinomial logistic regression model was performed to identify factors associated with chemotherapy receipt among all LN+ patients. Results: From 1/2016-3/2018, we identified 296 consecutive patients with HR+/HER2-/LN+ breast cancer. 200 (68%) patients had RS testing and 128 (64%) met reflex criteria. Reasons for not meeting RS reflex criteria included age > 65 (n = 35), grade III disease (n = 35) and N2/N3 tumors (n = 10). Among the 200 patients with RS, 122 (61%) had RS < 18, 67 (34%) had RS 18-30, and 11 (6%) had RS > 30. Only 68/200 (34%) patients with RS received chemotherapy as compared to 54/96 (56%) patients without RS (p = 0.0004). Compared to patients without RS testing, the odds of receiving chemotherapy were less with RS < 18 (OR = 0.46). The odds of receiving chemotherapy were greater with ≥3 positive LNs versus 1 positive LN (OR = 3.40). Conclusions: The majority of HR+/HER2-/LN+ patients undergoing upfront surgery at our center receive RS testing (200/296), with 122 (61%) resulting in low risk RS. Patients with low risk scores (RS < 18) were less likely to receive chemotherapy. While nodal involvement remains a common driver of chemotherapy use, our study demonstrates that RS testing provides clinically useful information in this population.

TPS1115 Background: A previous study from our group showed that approximately 9% of metastatic breast cancer (MBC) is hypermutated, defined as a tumor mutational burden (TMB) ≥10 Mutations/Megabase (Mut/Mb). The aim of this study is to evaluate if patients with hypermutated HER2-negative MBC benefit from the combination of nivolumab plus ipilimumab. Methods: This is an open-label, single-arm, multicenter, phase 2 study assessing the efficacy of nivolumab 3 mg/Kg intravenously (IV) every 14 days plus Ipilimumab 1 mg/Kg IV every 6 weeks in subjects with hypermutated metastatic HER2-negative breast cancer. Patients with measurable HER2-negative MBC, TMB ≥10 Mut/Mb assessed by a cancer-gene panel evaluating > 300 genes and performed in a CLIA-certified laboratory, and 0-3 prior lines of chemotherapy in the advanced setting are eligible. The primary objective is overall response rate according to RECIST 1.1. Secondary objectives include the safety and tolerability of the combination, progression-free survival, and overall survival. The study will follow a two-stage design. In the first stage 14 patients will be enrolled. If there is at least one patient with objective response, accrual will continue to the second stage where an additional 16 patients will be enrolled. If there are at least 4 patients with an objective response among the 30 patients, the regimen will be considered worthy of further study. If the true response rate is 5%, the chance the regimen is declared worthy of further study is less than 5%. If the true response rate is 25%, the chance that the regimen is declared worthy of further study is > 90%. Tumor biopsies, peripheral blood, and stool collection are mandatory and will be obtained at baseline, on treatment (end of cycle 1), and at disease progression and will be assessed for potential biomarkers of treatment response. The trial was activated in February 2019, and accrual should be completed in 18 months. Clinical trial information: NCT03789110.

502 Background: HER2 targeted therapy without chemotherapy may be insufficient to completely eradicate a HER2+ cancer in cases of significant intratumor HER2 heterogeneity (ITH-HER2). Methods: We conducted a single-arm phase II study enrolling centrally confirmed HER2+ breast cancer. Pts received 6 cycles of T-DM1 plus Pertuzumab before surgery. Central ITH-HER2 was assessed on baseline ultrasound-guided core biopsies from 2 distinct areas of each tumor (3 cores/site). ITH-HER2 was defined as at least one of the six areas demonstrating either 1) HER2 positivity by FISH in > 5% and < 50% of tumor cells (i.e., CAP guideline) or 2) an area of tumor that tested HER2 negative. The primary objective is the association between pathologic complete response (pCR) and ITH, stratified by ER status. pCR defined as residual cancer burden (RCB) 0. Results: 164 pts with centrally confirmed HER2+ tumors were enrolled from 1/2015 to 1/2018. 2 pts withdrew consent. Median tumor size by imaging was 2.8 cm (IQR 2.1-3.8cm); 111 (69%) were ER+ and 51 (32%) ER-. 8 pts discontinued tx (6 due to disease progression, 2 due to toxicity). 49% of pts had a pCR (RCB-0), 14% RCB-I, 26% RCB-II and 11% RCB-III. Higher rates of RCB-0 were seen in ER- (65%) versus ER+ (42%). ITH-HER2 was detected in 10% (16/157) of evaluable cases. No pCR was observed among cases classified as heterogeneous (RCB-I 25%, RCB-II 25%, RCB-III 50%). The study met its primary endpoint by demonstrating a significant association between ITH-HER2 and pCR stratified by ER status (p < .0001). Secondary analysis also demonstrated a significant association between ITH-HER2 and pathologic response defined as RCB 0 or I (OR = 5.6, p = 0.004). Exploratory analysis revealed higher rates of RCB-0 among tumors centrally classified as HER2 3+ (56% [66/118]) versus HER2 2+ (27% [10/37]), (OR = 3.4, p = 0.002). The association of ITH-HER2 and pCR was maintained when stratifying by ER status and HER2 IHC (2+ vs. 3+), (p = 0.002). Conclusions: ITH-HER2 assessed by routine pathology evaluation is a strong predictor of pCR to a dual-HER2 targeted therapy regimen. If validated, ITH-HER2 may need to be considered in selection of pts for HER2-targeted regimens without chemotherapy in the curative setting. Clinical trial information: NCT02326974.

1004 Background: Studies of checkpoint inhibitor monotherapy show only modest activity in HR+ MBC. We report data from the first randomized study comparing E plus P versus E alone in HR+/HER2- MBC. Methods: Eligible patients (pts) had HR+/HER2- MBC, ≥2 lines of hormonal therapies and 0-2 lines of chemotherapy for MBC. Pts were randomized 1:1 to E 1.4mg/m2 intravenously (IV) on d1 and d8 with P 200 mg/m2 IV on d1 of a 21-day cycle (Arm A) or E alone (Arm B). At time of progression, pts in arm B could crossover and receive P alone. Primary endpoint was progression-free survival (PFS). Key secondary endpoints were: objective response rate (ORR) and overall survival (OS). Exploratory analyses assessed the association between PFS and PD-L1 status, tumor-infiltrating lymphocytes (TILs), neutrophil-lymphocyte ratio (NLR), tumor mutation burden (TMB), and genomic alterations by next generation sequencing on archival tissue. Results: 88 pts initiated protocol therapy; the median age was 58, median prior lines of chemotherapy 1, prior lines of hormonal therapy 2. Median follow-up was 6.3 months. Median PFS and ORR were not different between Arms A and B (PFS 4.1 vs 4.2 months p = 0.38; ORR 25% and 34% respectively (p = 0.49). 14 patients initiated crossover treatment with pembrolizumab; 1 patient experienced a PR (ORR 7%). All-cause AEs occurred in 100% of pts (G3-4, 54.6%) including 2 treatment related deaths on Arm A, both from known AEs attributed to both drugs. PD-L1 assay was performed in 65 pts: 24 (36.9%) had PD-L1 positive ( > 1% with 22C3, centrally tested) tumors. PD-L1 status, TILs, NLR, TMB, and genomic alterations were not associated with PFS (Table). Updated data, including OS and genomic results, will be presented. Conclusions: Among pts with HR+/HER2- MBC, the combination of E and P was not associated with longer PFS than E alone in the ITT or PD-L1+ population, though the PD-L1+ subgroup had very limited power to assess P benefit. Clinical trial information: NCT03051659. [Table: see text]

1047 Background: RT is frequently used for palliation in MBC. In animal models its use has been reported to induce distant (abscopal) tumor responses when combined with immune checkpoint inhibitors. Here, we report the safety and efficacy of palliative RT plus pembrolizumab in a phase II single-arm study in patients (pts) with HR+/HER2- MBC. Methods: Eligible pts had HR+/HER2- MBC, ECOG PS 10%. Among 5 pts with PD-L1 status available, 2 were positive. Six pts had NLR > 4. Conclusions: Pembrolizumab combined with RT was well-tolerated, and no unexpected adverse events were observed; however, clinical benefit of the combination was not demonstrated in this heavily pretreated HR+ population. Clinical trial information: NCT03051672.

517 Background: Use of growth factors (GF) adds considerable expense and some toxicity to adjuvant breast cancer chemotherapy. We tested the feasibility and safety of omitting routine GF use during the T portion of DD AC-T. Methods: This is a prospective, single-arm study in which patients (pts) who completed 4 cycles of DD-AC proceeded to DD-T 175 mg/m2 every two weeks (wks) without routine GF (NCT02698891). Key inclusion: age≤ 65, ECOG PS≤1, absolute neutrophil count (ANC) ≥1500/mm3, and no febrile neutropenia (FN) during DD-AC. Criteria to treat for T included ANC ≥1000/mm3. Peg-F was given only if pts had FN in a prior cycle, or at investigator discretion if infection or treatment delay > 1 wk. Once Peg-F was given, pts received it in all future cycles. The primary endpoint was the rate of T completion ≤ 7 wks from cycle 1 day 1 (C1D1) to C4D1. Secondary endpoints included total use of Peg-F, rates of hematologic toxicity and FN, reasons for dose modification or hold. If ≥85% of pts completed T on time, the regimen would be considered feasible. If the true on-time completion rate is 75%, the chance the regimen would be declared infeasible is 91%, and if it is 85% the chance that the regimen is falsely declared infeasible is 10% (power = 0.899). ≥100/125 pts had to complete T on time for the regimen to be deemed successful. Results: Among 127 pts enrolled, 125 received ≥1 dose of protocol therapy and are included in the analysis. Median age at registration was 46 (range 21-65). Median C1D1 ANC was 7500/mm3 (range 1500-20500). 112 (90%) (95% CI 83-94%) pts completed DD-T ≤ 7 wks, and 3 (2%) completed within > 7 wks (2 due to neutropenia); 10 (8%) did not complete all cycles of T. Omission of Peg-F was not causally related to non-completion of T in any pts. The most common reasons for dose reduction or delays were non-hematologic. One pt had FN but was able to complete T on time. Eight (6.4%) pts received Peg-F during the trial. Conclusions: Omission of routine GF use during DD-T according to a pre-specified algorithm appears safe, feasible, and was associated with a 95.7% reduction in use of Peg-F, relative to the current standard of care. Additional analyses including cost implications are ongoing. Clinical trial information: NCT02698891.

2543Background: The hazard ratio is the most common measure of treatment effect in oncology trials, but many trials commonly violate the underlying Cox proportional hazards assumption. We aimed to ...

TPS2079 Background: Patient with glioblastoma (GBM) with unmethylated MGMT promoters derive limited benefit from temozolomide (TMZ) and have dismal outcome. Prioritizing the numerous available therapies and biomarkers for late stage testing requires an efficient clinical testing platform. INSIGhT (NCT02977780) is a biomarker-based, Bayesian adaptively randomized, multi-arm phase II platform screening trial for patients with newly diagnosed GBM and unmethylated MGMT promoters. Methods: INSIGhT compares experimental arms to a common control of standard concurrent TMZ and radiation therapy (RT) followed by adjuvant TMZ. The primary endpoint is overall survival (OS). Patients with newly diagnosed, unmethylated GBM that are IDH R132H mutation negative, and with genomic data available or who consent to whole exome sequencing through the ALLELE companion study for biomarker grouping are eligible. Two experimental arms consist of concurrent RT/TMZ followed by adjuvant neratinib (EGFR, HER2, and HER4 inhibitor) or abemaciclib CDK 4/6 inhibitor), respectively, in place of TMZ. The other experimental arm is CC-115 (TORC1/2 and DNA PK inhibitor), which replaces TMZ in both the concurrent and adjuvant phases. Biomarker groups include: EGFR + patients with EGFR amplification/mutation; PI3K + patients with PIK3CA mutation/amplification, PIK3R1 mutation, AKT3amplification, PIK3C2B > 1 copy gain, or PTEN dual loss; CDK: + patients with wild-type RB1 and CDK4 amplification, CDK6 amplification, or CDKN2A > 1 copy loss. Given the lack of pretrial biomarker data and the anticipated overlap of the groups, randomization will initially be equal. As the trial progresses, randomization probabilities will be adapted based on the Bayesian estimation of the probability of treatment impact on progression-free survival (PFS). These randomization probabilities can vary among the biomarker groups so predictive biomarkers will be identified and utilized if present. Treatment arms may drop due to low probability of treatment impact on OS, and new arms may be added. Experimental arms are compared only with control and should be thought of as discrete experimental questions, with INSIGhT being open to new investigators with proposed therapeutic hypotheses. Clinical trial information: NCT02977780.

e14057Background: Identifying predictive biomarkers is essential to precision medicine and can make therapeutic development more efficient. A critical decision in designing of biomarker-based clini...

e13009 Background: Multi-arm randomized controlled trials with Bayesian interim analyses improve efficiencies over conventional two arm trials. We expand on this to include biomarkers and a set of heterogeneous drugs that are linked to distinct treatment-biomarker interaction hypotheses. While biomarker hypotheses can be incorporated into clinical trials in a myriad of ways, we developed a novel methodology to incorporate prior information to gain the efficiency associated with biomarker-only trials while preserving the valuable biomarker information generated by all patient trials. Experimental arms are linked to biomarker-defined subgroups within the enrolled population, with treatment/biomarker links estimated with various degrees of uncertainty by previous studies including preclinical experiments. Methods: We propose statistical designs and analyses of multi arm trials using a priori information on treatment-biomarker interactions formalized through Bayesian modeling. A priori information is represen...

5531 Background: Numerous gene signatures of patient prognosis for late-stage, high-grade ovarian cancer have been published, but diverse data and methods have made these difficult to compare objec...

2005 Background: Bayesian-based trial design has the ability to utilize accumulating data in real time to alter the course of the trial, thereby enabling dynamic allocation to experimental arms and earlier dropping of ineffective arms. This flexibility results in a potentially more efficient trial framework by increasing the probability of enrollment to arms that show evidence of efficacy. In this study we considered a hypothetical scenario in which patients who have been previously treated on several separate experimental protocols at the Dana-Farber/Harvard Cancer Center and the University of California, Los Angeles were instead enrolled in a single multi-arm protocol utilizing a Bayesian adaptively randomized trial design. The purpose was to determine whether similar scientific results could have been accomplished more efficiently. Methods: We generated an adaptive randomization procedure that was retrospectively applied to primary patient data from four separate phase II clinical trials in patients with recurrent glioblastoma. We then compared AR to more conventional trial designs using realistic hypothetical scenarios consistent with survival data reported in the literature. Our primary endpoint is the number of patients needed to achieve a desired statistical power. Results: If our phase II trials had been a single multi-arm AR trial, bevacizumab would have been identified as an efficacious therapy, and 30 fewer patients would have been needed compared to a multi-arm balanced randomized (BR) design. More generally, Bayesian AR trial design for patients with glioblastoma would result in trials with fewer overall patients with no loss in statistical power, and in more patients randomized to effective treatment arms. For a trial with a control arm, two ineffective arms and one effective arm with hazard ratio 0.6, a median of 47 patients would be randomized to the effective arm compared with 35 in a BR design. Conclusions: Given the desire for control arms in phase II trials, an increasing number of experimental therapeutics for patients with glioblastoma, and a relatively short time for events, Bayesian adaptive designs are attractive for clinical trials in glioblastoma.