National Cancer Institute Molecular Analysis For Therapy Choice Research Articles

NCI MATCH (EAY131) arm Z1D: Retrospective evaluation of attributions of adverse events experienced on nivolumab in patients with MSI-high tumors.

e14711 Background: Machine learning algorithms for prediction of Adverse Events (AEs) due to immune checkpoint inhibitors (ICIs) rely on clinically annotated data to serve as gold standards. Clinical trials are the best sources of such data, but audits to assess validity are rare, compromising future machine learning efforts. We reviewed AEs reported on the National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) EAY-131 Arm Z1D. Patients matched to Arm Z1D due to MSI-high status were treated with nivolumab. Routine and serious AEs were reported to NCI. Attribution data from AEs on 29 patients with tumor genomics suitable for machine learning approaches were obtained and reviewed retrospectively. Methods: Documented AEs were sorted by grade via the Common Terminology for Adverse Events (CTCAE) system and categorized as related or unrelated to ICIs. Primary source data reported to NCI were reviewed. If any ambiguity was found, trial sites were queried to provide additional information and documentation. Decisions were made by an experienced NCI medical officer with the benefit of additional data and site responses. All serious AEs and a subset of ambiguous routine AEs were assessed to determine if attribution conformed to expected attribution per the nivolumab package insert. Uncertain attribution required additional queries, including email or phone correspondence. Grade 1 or 2 events were only evaluated to determine if any attribution to ICIs was possible. If a higher grade event was attributed to ICIs, there was no further query for lower grade events on the same patient. Results: A total of 639 Grade 1, 152 Grade 2, 75 Grade 3, 10 Grade 4, and 5 Grade 5 events were reported among the 29 patients. One Grade 5 event, 5 Grade 4 events, 13 Grade 3 Events, 9 Grade 2 events, and 6 Grade 1 events were considered less likely to be attributed to ICIs compared with the initial sponsor or site investigator assessment of possibly, probably, or definitely related. Two Grade 4 events, 1 Grade 2 event, and 8 Grade 1 events were considered to be possibly, probably, or definitely due to ICIs compared with the initial sponsor or site assessment of less likely or unrelated. Site responsiveness to retrospective queries from the sponsor varied. Additional lab values, primary source documentation, and communication with site investigators were necessary to make modifications. A total of 11 AEs required additional queries to site investigators. Conclusions: Retrospective reviews of real-time site and sponsor assessments of attribution exhibit significant modification of initial attribution to ICIs in this review of 29 patients on Arm Z1D of the NCI MATCH trial. Machine learning approaches for prediction of AEs on patients treated with ICIs may require careful assessments of primary source data to develop adequate training sets, given the paucity of gold standard data available.

Molecular profiling of ctDNA from NCI-MATCH patients enrolled for treatment with mTOR1/2 inhibitor sapanisertib (arm M) and the Hedgehog pathway inhibitor vismodegib (arm T).

3047 Background: The National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) multi-arm phase II clinical trial tested biopsies from patients with advanced refractory cancer to assign treatment based on tumor molecular profile. Enrollment criteria for treatment with the mTOR1/2 inhibitor sapanisertib in Arm M, or the Hedgehog pathway inhibitor vismodegib in Arm T, included detection of mutations in TSC1/TSC2 or PTCH1/SMO in tissue, respectively. For a subset of patients enrolled on these two arms, matched plasma was also collected at time of enrollment or on treatment to evaluate ctDNA. Methods: Screening for NCI-MATCH was based on tissue sequencing that was initially performed centrally using the 143-gene Oncomine Comprehensive Assay version 2 (OCAv2) and later referred to NCI-MATCH Designated Laboratory (DL) Network to perform sequencing for patient referral. Cell-free DNA (cfDNA) extracted from plasma collected in Streck tubes was input at 10 – 30 ng into a modified version of the Illumina TruSight Oncology 500 assay (TSO 500). Following error correction and read collapsing post-sequencing, variant calls were made and concordance between cfDNA libraries meeting prespecified QC criteria and tissue was evaluated. Results: Plasma, representing multiple collection timepoints, was available for 30/49 Arm M and 21/34 Arm T enrolled patients. A total of 95 cfDNA libraries, including replicates, were constructed and sequenced. Unique patient samples were tested across various timepoints (n = 35 pretreatment; n = 35 Cycle 2 Day 1; n = 8 at progression) plus replicates, with 90 (94.7%) libraries passing all prespecified QC criteria. Among Arm M patients there were 3 patients whose enrollment variant did not overlap with the TSO 500 gene panel. In total, 14 of 27 (51.8%) patients in Arm M and 18 of 21 (85.7%) patients in Arm T were concordant for enrollment mutation for at least one collection timepoint. In Arm T, 85.7% (n=38) of patient samples had at least one oncogenic or likely oncogenic (O/LO) mutation in the tumor sample at enrollment, with 73.6% (n = 28) of those also being detected by cfDNA. Interestingly, within Arm M and Arm T, a total of 7 patient samples were detected as MSI-H, with corresponding bTMB scores ranging from ~6 to > 100 Mut/Mb (median = 86.7). Conclusions: Detection of the enrollment mutations in treatment arms was observed in 51.8% - 85.7% of patients tested by ctDNA with overall variant concordance between tissue and ctDNA of at least 73.6%. These findings support the use of liquid biopsy as a tool for understanding genomic profiles of cancer patients both at diagnosis and progression, less invasively than standard tissue biopsies. In addition, 7 patients were identified as MSI-H, suggesting that blood-based testing could complement tissue testing for identifying patients who may benefit from targeted therapy.

Progress report of a cross-organ and biomarker-based platform study: BELIEVE trial NCCH1901.

TPS1611 Background: Cancer Genomic Medicine (CGM) has been developing and more patients have received the comprehensive genomic profiling (GCP) test under national healthcare system in Japan. However, the number of patients who could receive the genomically matched therapy based on the results of CGP tests was limited. To improve the drug accessibility, we initiated a cross-organ, biomarker-based clinical trial (NCCH1901, the BELIEVE trial). It was a platform trial and could be considered as the Japanese counterpart to the National Cancer Institute's Molecular Analysis for Therapy Choice (NCI-MATCH) study in the United States. Methods: It was a multi-institutional joint research nationwide led by NCCH with a master’s protocol. Several drug cohorts were pre-established, allowing patients to participate in molecularly targeted treatments recommended as a result of CGP test. All drugs were provided for free from the pharmaceutical companies. The objective of this trial was to administer off-label drugs for the respective genomic abnormalities and collect efficacy and safety data for patients of the drugs. The primary endpoint was the response rate based on the best overall response for up to 16 weeks. This trial has been conducted under both of the Japanese Clinical Trial Act and the patient-proposed healthcare services which enable patients to receive unapproved, off-label drugs partially covered by National Health Insurance. The protocol and the informed consent forms were approved by the National Cancer Center Hospital (NCCH) certified review board and the Ministry of Health, Labour and Welfare (MHLW). The trial was registered in the Japan Registry of Clinical Trials (jRCTs031190104). As of January 31, 2023, we have 20 drug cohorts and the planned number of enrollments for each cohort was 50 with the target lesions and 30 without the target lesions. There were more than 500 patients so far. The most common cancer was brain tumours, followed by carcinoma of endocrine organs and the colorectal cancer. BRAF mutation and ERBB2 amplification were the frequent genomic abnormalities registered in this trial. Five drug cohorts have already finished recruiting and 2 drug cohorts assessed their efficacies. We are still continuing to discuss several pharmaceutical companies to increase the drugs cohorts in this platform. This trial is now recruiting the patients and is contributing significantly to increase the treatment opportunity for the cancer patients. Clinical trial information: jRCTs031190104 .

235 (PB115) - Phase II trial of afatinib in patients with EGFR-mutated solid tumors excluding lung cancer: results from the NCI-MATCH ECOG-ACRIN Trial (EAY131) Subprotocol A

Background: Molecularly targeted treatment with epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) have prolonged survival of pts with activating EGFR mutations in non-small cell lung cancer (NSCLC). Afatinib is an ErbB family blocker that demonstrated clinical benefit in pts with NSCLC with common and uncommon activating EGFR mutations. The activity of afatinib in other tumor types with EGFR mutations is less well known. National Cancer Institute-Molecular Analysis for Therapy Choice (NCI-MATCH) Arm EAY131-A evaluated afatinib in tumors beyond lung cancer that harbored EGFR activating alterations.

Blood-based detection of actionable alterations from NCI-MATCH patients with no tissue results.

3035 Background: The National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) multi-arm phase II clinical trial tested tumor tissue from 5,954 patients with advanced refractory cancer to assign treatment based on the molecular profile. Molecular profiling was successful for 93% of patients. For 267 of the patients who were not enrolled because molecular profiling was not successful, plasma cfDNA was evaluated to provide insight into the potential utility of blood-based testing in a broad spectrum of histologies when tissue is not evaluable. Methods: Cell-free DNA was extracted from plasma collected from Streck blood tubes and quantitated. Libraries were constructed using ³ 15 ng cfDNA into the Illumina TruSight Oncology 500 ctDNA RUO Assay, including unique molecular identifiers and duplex barcodes for error correction. Libraries were sequenced on the NovaSeq 6000 with S4 XP flow cells. Results: Of the 267 samples, 250 samples (94%) were evaluable, representing 72 histologies, including colorectal cancer (N = 36), lung adenocarcinoma (N = 15), pancreatic adenocarcinoma (N = 14), and invasive breast carcinoma (N = 12). Of these, 231 (92%) had ³ 1 OncoKB annotated mutation, with 208 patients (83%) having putative somatic mutations detected in genes not commonly associated with clonal hematopoiesis. The most common somatic mutations were in TP53, KRAS, APC, and PIK3CA, reported in 51%, 20%, 12%, and 12% of patients respectively. A total of 109 patients (44%) had ³ 1 actionable mutation of interest (aMOI) reported that could have been used for treatment assignment in the NCI-MATCH clinical trial. After applying histology and molecular exclusions, 75 patients (30%) had ³ 1 aMOI. The most common assignable treatment arms were Z1B/Z1BX1 (palbociclib with CCND1/2/3, N = 13), Z1F (copanlisib with PIK3CA Mutations, N = 13), S1/S1X1 (trametinib with NF1 mutation, N = 12), and Z1C/Z1CX1 (palbociclib with CDK4/CDK6 Amplification and Rb Expression by IHC, N = 10). Mutations in genes commonly associated with clonal hematopoiesis (CH) were prevalent in this population. Along with the expected high frequency of DNMT3A (21% of patients) and TET2 (11%) mutations, PPM1D mutations were the highest amongst CH genes, with 61 patients (24%) having ³ 1 PPM1D mutation, likely due to the heavily pre-treated nature of these patients. Conclusions: Variants observed in the blood are consistent with what is reported in the tissue. Using liquid biopsy when tissue is not evaluable can expand the ability of patients to obtain mutation information that can inform treatment compared to using tumor tissue only. Cell-free DNA provided valuable mutation information for these patients and could have resulted in up to an additional 75 patients being eligible for treatment selection based on their mutation profile. These results indicate that blood-based screening could be a tool for future NCI-sponsored clinical studies.

Effect of Capivasertib in Patients With an AKT1 E17K-Mutated Tumor

In the National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) trial, agents targeting genetic tumor abnormalities are administered to patients. In the NCI-MATCH subprotocol EAY131-Y trial, patients with an AKT1 E17K-mutated metastatic tumor received the pan-AKT inhibitor capivasertib. To assess the objective response rate (ORR) of capivasertib in patients with an AKT1 E17K-mutated tumor. Between July 13, 2016, and August 10, 2017, patients in the NCI-MATCH trial were enrolled and assigned to the subprotocol EAY131-Y nonrandomized trial. Patients included adults with an AKT1 E17K-mutated metastatic tumor that had progressed with standard treatment, and these patients were assigned to receive capivasertib. Tumor assessments were repeated every 2 cycles. Data analysis of this evaluable population was performed from November 8, 2019, to March 12, 2020. The study treatment was capivasertib, 480 mg, orally twice daily for 4 days on and 3 days off weekly in 28-day cycles until disease progression or unacceptable toxic effect. If patients continued hormone therapy for metastatic breast cancer, the capivasertib dose was 400 mg. The primary end point was the ORR (ie, complete response [CR] and partial response) according to the Response Evaluation Criteria in Solid Tumors criteria, version 1.1. Secondary end points included progression-free survival (PFS), 6-month PFS, overall survival, and safety. In total, 35 evaluable and analyzable patients were included, of whom 30 were women (86%), and the median (range) age was 61 (32-73) years. The most prevalent cancers were breast (18 [51%]), including 15 patients with hormone receptor (HR)-positive/ERBB2-negative and 3 with triple-negative disease, and gynecologic (11 [31%]) cancers. The ORR rate was 28.6% (95% CI, 15%-46%). One patient with endometrioid endometrial adenocarcinoma achieved a CR and remained on therapy at 35.6 months. Patients with confirmed partial response had the following tumor types: 7 had HR-positive/ERBB2-negative breast cancer, 1 had uterine leiomyosarcoma, and 1 had oncocytic parotid gland carcinoma and continued receiving treatment at 28.8 months. Sixteen patients (46%) had stable disease as the best response, 2 (6%) had progressive disease, and 7 (20%) were not evaluable. With a median follow-up of 28.4 months, the overall 6-month PFS rate was 50% (95% CI, 35%-71%). Capivasertib was discontinued because of adverse events in 11 of 35 patients (31%). Grade 3 treatment-related adverse events included hyperglycemia (8 [23%]) and rash (4 [11%]). One grade 4 hyperglycemic adverse event was reported. This nonrandomized trial found that, in patients with an AKT1 E17K-mutated tumor treated with capivasertib, a clinically significant ORR was achieved, including 1 CR. Clinically meaningful activity with single-agent capivasertib was demonstrated in refractory malignant neoplasms, including rare cancers. ClinicalTrials.gov Identifier: NCT00700882.

Tumor Genomic Profiling Practices and Perceptions: A Survey of Physicians Participating in the NCI-MATCH Trial.

To identify factors that may influence physician participation in tumor profiling studies and to assess the routine use of tumor profiling in clinical practice. Physicians in the National Cancer Institute-Molecular Analysis for Therapy Choice (NCI-MATCH) were invited to participate in an electronic survey consisting of 73 questions related to participation in genomic profiling studies, tumor profiling practices and education during usual patient care, and physician background and practice characteristics. The survey response rate was 8.9% (171 surveys returned of 1,931 sent). A majority of respondents practiced in academic medical centers (AMCs). Participation in NCI-MATCH increased workload and cost but resulted in increased professional satisfaction, confidence in treatment recommendation, and subsequent use of tumor profiling. Barriers to patient participation included length of wait time for results and lack of a therapeutic option from the testing. Physicians who worked in AMCs reported a higher use of tumor profiling than did those who worked in non-AMC settings (43% v 18%; P = .0009). Access to a molecular tumor board was perceived as valuable by 56%. The study identified a need for educational materials to guide both physicians and patients in the field of genomic profiling. Physicians who participate in NCI-MATCH perceive value to patient treatment that outweighs the additional effort required; survey results help identify barriers that may limit participation. The current findings have implications for the design of future genomic and other profiling studies.

Molecular Landscape and Actionable Alterations in a Genomically Guided Cancer Clinical Trial: National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH)

PURPOSETherapeutically actionable molecular alterations are widely distributed across cancer types. The National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) trial was designed to evaluate targeted therapy antitumor activity in underexplored cancer types. Tumor biopsy specimens were analyzed centrally with next-generation sequencing (NGS) in a master screening protocol. Patients with a tumor molecular alteration addressed by a targeted treatment lacking established efficacy in that tumor type were assigned to 1 of 30 treatments in parallel, single-arm, phase II subprotocols.PATIENTS AND METHODSTumor biopsy specimens from 5,954 patients with refractory malignancies at 1,117 accrual sites were analyzed centrally with NGS and selected immunohistochemistry in a master screening protocol. The treatment-assignment rate to treatment arms was assessed. Molecular alterations in seven tumors profiled in both NCI-MATCH trial and The Cancer Genome Atlas (TCGA) of primary tumors were compared.RESULTSMolecular profiling was successful in 93.0% of specimens. An actionable alteration was found in 37.6%. After applying clinical and molecular exclusion criteria, 17.8% were assigned (26.4% could have been assigned if all subprotocols were available simultaneously). Eleven subprotocols reached their accrual goal as of this report. Actionability rates differed among histologies (eg, > 35% for urothelial cancers and < 6% for pancreatic and small-cell lung cancer). Multiple actionable or resistance-conferring tumor mutations were seen in 11.9% and 71.3% of specimens, respectively. Known resistance mutations to targeted therapies were numerically more frequent in NCI-MATCH than TCGA tumors, but not markedly so.CONCLUSIONWe demonstrated feasibility of screening large numbers of patients at numerous accruing sites in a complex trial to test investigational therapies for moderately frequent molecular targets. Co-occurring resistance mutations were common and endorse investigation of combination targeted-therapy regimens.

Performance Characteristics of a Targeted Sequencing Platform for Simultaneous Detection of Single Nucleotide Variants, Insertions/Deletions, Copy Number Alterations, and Gene Fusions in Cancer Genome.

An increasing number of molecular laboratories are implementing next-generation sequencing platforms to identify clinically actionable and relevant genomic alterations for precision oncology. To describe the validation studies as per New York State-Department of Health (NYS-DOH) guidelines for the Oncomine Comprehensive Panel v2, which was originally tailored to the National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) trial. Accuracy, precision, and reproducibility were investigated by using 130 DNA and 18 RNA samples from cytology cell blocks; formalin-fixed, paraffin-embedded tissues; and frozen samples. Analytic sensitivity and specificity were tested by using ATCC and HapMap cell lines. High accuracy and precision/reproducibility were observed for single nucleotide variants and insertion/deletions. We also share our experience in the detection of gene fusions and copy number alterations from an amplicon-based sequencing platform. After sequencing analysis, variant annotation and report generation were performed by using the institutional knowledgebase. This study serves as an example for validating a comprehensive targeted next-generation sequencing assay with both DNASeq and RNASeq components for NYS-DOH.

Abstract A079: National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH): A successful precision medicine signal-seeking trial in patients (pts) with rare variants and refractory malignancies

Abstract Introduction: NCI-MATCH, developed by ECOG-ACRIN &amp; NCI, is the largest precision medicine study for pts with refractory malignancy. Over 1100 clinical sites in the National Clinical Trials Network enrolled pts. The purpose of the study is to identify potentially beneficial targeted treatments across tumor types with similar molecular abnormalities. Methods: The NCI Central IRB approved NCI-MATCH. Pts with refractory/no treatment available solid tumors, lymphomas or myelomas had a fresh biopsy profiled by next generation sequencing (143 genes, &amp;gt; 4000 single nucleotide variants, indels, amplifications &amp; targeted fusions). Pts are assigned by a defined algorithm to treatments with evidence of activity against tumors with the relevant molecular alteration. Pts are excluded if a treatment is FDA approved or known to be ineffective for their malignancy. After successfully sequencing fresh biopsies from 5540 pts, subprotocols with extremely rare variants lacked sufficient accrual. To address actionable variants with a prevalence of &amp;lt; 1.5%, we decided to accept clinical sequencing results from 30 commercial and academic laboratories vetted by NCI-MATCH to address relevant variants. These labs notify clinicians participating in NCI-MATCH if their pt’s tumor contains an actionable variant. Treatment continues until tumors became refractory, pt intolerance or withdrawal of consent. An objective response rate (ORR by RECIST) of &amp;gt; 16% among 31 eligible patients is considered a positive signal. Results: After screening 5540 pts, 37.6% had an actionable variant. After histology and treatment-specific exclusions, 17.8% were assigned and 69.5% enrolled on the assigned subprotocol. 11 of the initial 30 subprotocols reached completion with adequate follow-up. Of the first 11 evaluable subprotocols, 3 addressing rare variants had a positive signal: Nivolumab in pts with loss of expression of MLH1 or MSH2 (ORR 36%), capivasertib in pts with AKT mutations (ORR 23%), and dabrafenib + trametinib in pts with BRAF V600 mutations (ORR 33%). These molecular variants were found in 2%, 1.2% and 1.9% respectively, of screened pts. Two other subprotcocols (afatinib in ERBB2 mutations and AZD4547 in FGFR abnormalities) showed responses in rare tumors or specific variant subsets, respectively. As of July 15, 2019, an additional 378 of 432 (88%) pts have been assigned to a treatment with a clinical sequencing assay; 83% of these pts enrolled to 1 of 24 subprotocols, allowing completion of an additional 9 of the original 30 subprotocols and complete accrual to 2 of 5 recently added subprotocols. Four of 35 subprotocols closed for lack of accrual, 10 continue accruing and 4 are planned. Conclusions: Platform precision medicine trials can identify potentially useful targeted treatments for diverse malignancies in pts with uncommon tumors &amp; rare actionable variants, an unmet need. In a population of pts with refractory cancers, lymphomas and myelomas, 30-40% will have an actionable variant for targeted treatment (investigational or standard). Of the first 11 subprotocols with adequate follow-up, 3 (27%) showed a positive signal and an additional 2 showed responses in rare tumors or in a molecular subset, suggesting that the NCI-MATCH trial approach identifies useful targets for further exploration. Citation Format: Lyndsay N Harris, Robert J Gray, Barbara A Conley, Alice P Chen, Keith T Flaherty, Stanley R Hamilton, Paul M Williams, Chris Karlovich, David Patton, Shuli Li, Lisa M McShane, Larry V Rubinstein, Edith P Mitchell, James V Tricoli, Richard F Little, Carlos L Arteaga, Peter J O'Dwyer, NCI-MATCH team. National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH): A successful precision medicine signal-seeking trial in patients (pts) with rare variants and refractory malignancies [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr A079. doi:10.1158/1535-7163.TARG-19-A079

Analytical Validation of the Next-Generation Sequencing Assay for a Nationwide Signal-Finding Clinical Trial: Molecular Analysis for Therapy Choice Clinical Trial

The National Cancer Institute-Molecular Analysis for Therapy Choice (NCI-MATCH) trial is a national signal-finding precision medicine study that relies on genomic assays to screen and enroll patients with relapsed or refractory cancer after standard treatments. We report the analytical validation processes for the next-generation sequencing (NGS) assay that was tailored for regulatory compliant use in the trial. The Oncomine Cancer Panel assay and the Personal Genome Machine were used in four networked laboratories accredited for the Clinical Laboratory Improvement Amendments. Using formalin-fixed paraffin-embedded clinical specimens and cell lines, we found that the assay achieved overall sensitivity of 96.98% for 265 known mutations and 99.99% specificity. High reproducibility in detecting all reportable variants was observed, with a 99.99% mean interoperator pairwise concordance across the four laboratories. The limit of detection for each variant type was 2.8% for single-nucleotide variants, 10.5% for insertion/deletions, 6.8% for large insertion/deletions (gap≥4 bp), and four copies for gene amplification. The assay system from biopsy collection through reporting was tested and found to be fully fit for purpose. Our results indicate that the NCI-MATCH NGS assay met the criteria for the intended clinical use and that high reproducibility of a complex NGS assay is achievable across multiple clinical laboratories. Our validation approaches can serve as a template for development and validation of other NGS assays for precision medicine.