ENABLE-SG (Educate, Nurture, Advise, Before Life Ends for Singapore) as a proactive palliative care model: protocol for a hybrid type 1 effectiveness-implementation randomized wait-list controlled trial.

BACKGROUND: Specialist palliative care is often provided late in the patient's disease trajectory in response to uncontrolled symptoms. Shifting from this reactionary illness-stress paradigm to a proactive health-wellness approach, the ENABLE (Educate, Nurture, Advise, Before Life Ends) telehealth model aims to enhance the coping, stress and symptom management, self-care, and advance care planning skills of patients with advanced cancers and their caregivers. The ENABLE model has been culturally adapted to Singapore (ENABLE-SG) and pilot-tested. A hybrid type 1 effectiveness-implementation design will be used to evaluate the effectiveness of ENABLE-SG while collecting real-world implementation data. METHODS: This single-centre, assessor-blind, wait-list (immediately vs. 6 months) randomized controlled trial will recruit 300 adult patients within 60 days of an advanced cancer diagnosis and their family caregivers from the National Cancer Centre of Singapore. ENABLE-SG comprises structured psychoeducational sessions with a telehealth coach, covering essential topics of early palliative care. Participants will be assessed at baseline and every 3 months until patient's death, 12 months (caregivers), or end of study (patients). The primary outcome is patient quality of life 6 months after baseline. Secondary patient-reported outcomes include mood, coping, palliative care concerns, and health status. Secondary caregiver-reported outcomes include caregiver quality of life, mood, coping, and care satisfaction. Mixed-effects regression modelling for repeated measurements will be used. To assess the effectiveness of ENABLE-SG versus usual care, patient and caregiver outcomes at 6 months will be compared. To compare earlier versus delayed ENABLE-SG, patient and caregiver outcomes at 12 months will be compared. Within the hybrid type 1 effectiveness-implementation design, implementation outcomes will be evaluated in both the early and delayed groups. Acceptability, adoption, appropriateness, and feasibility will be assessed using a feedback survey and semi-structured interviews with a purposive sample of patients, caregivers, and healthcare providers. Transcribed interviews will be analysed thematically. Other implementation outcomes of penetration, fidelity, and cost will be assessed using records of study-related processes and summarized using descriptive statistics. A cost-effectiveness analysis will also be conducted. DISCUSSION: This study will assess both effectiveness and implementation of ENABLE-SG. Insights into implementation processes can facilitate model expansion and upscaling. TRIAL REGISTRATION: Registered prospectively on ClinicalTrials.gov, NCT06044441. Registered on 21/09/2023.

Addressing cellular heterogeneity in tumor and circulation for refined prognostication.

Despite pronounced genomic and transcriptomic heterogeneity in non-small-cell lung cancer (NSCLC) not only between tumors, but also within a tumor, validation of clinically relevant gene signatures for prognostication has relied upon single-tissue samples, including 2 commercially available multigene tests (MGTs). Here we report an unanticipated impact of intratumor heterogeneity (ITH) on risk prediction of recurrence in NSCLC, underscoring the need for a better genomic strategy to refine prognostication. By leveraging label-free, inertial-focusing microfluidic approaches in retrieving circulating tumor cells (CTCs) at single-cell resolution, we further identified specific gene signatures with distinct expression profiles in CTCs from patients with differing metastatic potential. Notably, a refined prognostic risk model that reconciles the level of ITH and CTC-derived gene expression data outperformed the initial classifier in predicting recurrence-free survival (RFS). We propose tailored approaches to providing reliable risk estimates while accounting for ITH-driven variance in NSCLC.

Concurrent Single-Cell RNA and Targeted DNA Sequencing on an Automated Platform for Comeasurement of Genomic and Transcriptomic Signatures.

BACKGROUND: The comeasurement of both genomic and transcriptomic signatures in single cells is of fundamental importance to accurately assess how the genetic information correlates with the transcriptomic phenotype. However, existing technologies have low throughput and laborious work flows. METHODS: We developed a new method for concurrent sequencing of the transcriptome and targeted genomic regions (CORTAD-seq) within the same single cell on an automated microfluidic platform. The method was compatible with the downstream library preparation, allowing easy integration into existing next-generation sequencing work flows. We incorporated a single-cell bioinformatics pipeline for transcriptome and mutation analysis. RESULTS: As proof of principle, we applied CORTAD-seq to lung cancer cell lines to dissect the cellular consequences of mutations that result in resistance to targeted therapy. We obtained a mean detection of 6000 expressed genes and an exonic rate of 50%. The targeted DNA-sequencing data achieved a 97.8% detection rate for mutations and allowed for the identification of copy number variations and haplotype construction. We detected expression signatures of tyrosine kinase inhibitor (TKI) resistance, epidermal growth factor receptor (EGFR) amplification, and expansion of the T790M mutation among resistant cells. We also identified characteristics for TKI resistance that were independent of EGFR T790M, indicating that other alterations are required for resistance in this context. CONCLUSIONS: CORTAD-seq allows assessment of the interconnection between genetic and transcriptomic changes in single cells. It is operated on an automated, commercially available single-cell isolation platform, making its implementation straightforward.

Effects of Single Nucleotide Polymorphisms on the Binding of Afatinib to EGFR: A Potential Patient Stratification Factor Revealed by Modeling Studies.

The anticancer drug afatinib has been found to be more effective at inhibiting the oncogenic EGFR mutant exon 19 deletion (19del) over the oncogenic EGFR mutant L858R. The underlying mechanism has been hypothesized to result from differences in structural constraints introduced by the mutations and stabilizing interactions afforded by a buried water molecule in 19del (Kannan, S.; et al. Sci. Rep. 2017, 7, 1540). The COSMIC cancer database is mined for EGFR sequences to discover that several mutations in the form of Single nucleotide polymorphisms (SNPs) line this hydration cavity. In this work, the effects of these SNPs on the affinity of afatinib for EGFRWT and oncogenic mutants EGFRL858R and EGFR19del were studied using free energy perturbation and thermodynamic Integration calculations. The simulations reveal that several SNPs have significant effects on the affinity of afatinib for the mutant EGFRs carrying the SNPs and may thus have clinical implications relating to emergence of resistance to afatinib, thus potentially impacting the choice of EGFR inhibitors in the clinic.

Mutational Aberrations Detected in Mucinous Epithelial Ovarian Cancer of Asian Women.

BACKGROUND: Mucinous epithelial ovarian cancers (mEOCs) respond poorly to conventional chemotherapy and have a poor prognosis in advanced stages. The genomic landscape for mEOC in the Asian settings is ill defined. We seek to identify various mutational aberrations present in mEOC and correlate them with clinical outcomes. METHODS: A total of 199 cases of mEOC were identified from a prospectively maintained gynecologic oncology tumor database. DNA was extracted and analyzed for KRAS mutations by using Sanger sequencing. Further MassArray sequencing was performed on 45 samples. Clinicopathologic correlation was performed with the results obtained. FINDINGS: KRAS mutation status was evaluable in 124 cases. Fifty-five percent (68/124) were KRAS negative, whereas 45% (56/124) harbored a KRAS mutation, lower than that in Western populations. Successful ascertainment of both KRAS and HER2 statuses by Sanger sequencing occurred for 105 cases. The proportion of the double-positive subtype (HER2+ and KRAS positive) was 8% (8/105); double-negative subtype (HER2- and KRAS negative), 34% (36/105); and cases with mutation in either KRAS or HER2, 58% (61/105). The KRAS mutation rate was 44%, 50%, and 29% among Chinese, Indians, and Malays, respectively. There was no significant difference in overall survival (P = 0.952) or progression-free survival (P = 0.635) between KRAS-positive and KRAS-negative patients. Similar results were observed for progression-free survival (P = 0.206) and overall survival (P = 0.440) when outcomes were examined between the 4 groups based on KRAS and HER2 mutation. Patients in the double-negative mutation subgroup had higher risk for death/progression compared with patients in the other 3 mutation subgroups. Further MassARRAY multiplexed profiling was performed in patients with sufficient DNA material (n = 45) and yielded KRAS mutations (n = 16), PDGFRA mutations (n = 3), PIK3CA (n = 1) and KIT (n = 1), and HRAS, FGFR, MET, and NRAS (n = 1 each). CONCLUSIONS: Our study provides further knowledge about the mutational aberrations in mEOC in Asian populations. Neither the presence of KRAS mutation nor their correlation with HER2 mutations influenced outcomes.

Single-cell profiling approaches to probing tumor heterogeneity.

Tumor heterogeneity is a major hindrance in cancer classification, diagnosis and treatment. Recent technological advances have begun to reveal the true extent of its heterogeneity. Single-cell analysis (SCA) is emerging as an important approach to detect variations in morphology, genetic or proteomic expression. In this review, we revisit the issue of inter- and intra-tumor heterogeneity, and list various modes of SCA techniques (cell-based, nucleic acid-based, protein-based, metabolite-based and lipid-based) presently used for cancer characterization. We further discuss the advantages of SCA over pooled cell analysis, as well as the limitations of conventional techniques. Emerging trends, such as high-throughput sequencing, are also mentioned as improved means for cancer profiling. Collectively, these applications have the potential for breakthroughs in cancer treatment.

An ultra-high-throughput spiral microfluidic biochip for the enrichment of circulating tumor cells.

The detection and characterization of rare circulating tumor cells (CTCs) from the blood of cancer patients can potentially provide critical insights into tumor biology and hold great promise for cancer management. The ability to collect a large number of viable CTCs for various downstream assays such as quantitative measurements of specific biomarkers or targeted somatic mutation analysis is increasingly important in medical oncology. Here, we present a simple yet reliable microfluidic device for the ultra-high-throughput, label-free, size-based isolation of CTCs from clinically relevant blood volumes. The fast processing time of the technique (7.5 mL blood in less than 10 min) and the ability to collect more CTCs from larger blood volumes lends itself to a broad range of potential genomic and transcriptomic applications. A critical advantage of this protocol is the ability to return all fractions of blood (i.e., plasma (centrifugation), CTCs and white blood cells (WBCs) (size-based sorting)) that can be utilized for diverse biomarker studies or time-sensitive molecular assays such as RT-PCR. The clinical use of this biochip was demonstrated by detecting CTCs from 100% (10/10) of blood samples collected from patients with advanced-stage metastatic breast and lung cancers. The CTC recovery rate ranged from 20 to 135 CTCs mL(-1) and obtained under high purity (of 1 CTC out of every 30-100 WBCs which gives ∼4 log depletion of WBCs). They were identified with immunofluorescence assays (pan-cytokeratin+/CD45-) and molecular probes such as HER2/neu.

Use of Natural Language Processing to Infer Sites of Metastatic Disease From Radiology Reports at Scale.

PURPOSE: To evaluate natural language processing (NLP) methods to infer metastatic sites from radiology reports. METHODS: A set of 4,522 computed tomography (CT) reports of 550 patients with 14 types of cancer was used to fine-tune four clinical large language models (LLMs) for multilabel classification of metastatic sites. We also developed an NLP information extraction (IE) system (on the basis of named entity recognition, assertion status detection, and relation extraction) for comparison. Model performances were measured by F1 scores on test and three external validation sets. The best model was used to facilitate analysis of metastatic frequencies in a cohort study of 6,555 patients with 53,838 CT reports. RESULTS: The RadBERT, BioBERT, GatorTron-base, and GatorTron-medium LLMs achieved F1 scores of 0.84, 0.87, 0.89, and 0.91, respectively, on the test set. The IE system performed best, achieving an F1 score of 0.93. F1 scores of the IE system by individual cancer type ranged from 0.89 to 0.96. The IE system attained F1 scores of 0.89, 0.83, and 0.81, respectively, on external validation sets including additional cancer types, positron emission tomography-CT ,and magnetic resonance imaging scans, respectively. In our cohort study, we found that for colorectal cancer, liver-only metastases were higher in de novo stage IV versus recurrent patients (29.7% v 12.2%; P < .001). Conversely, lung-only metastases were more frequent in recurrent versus de novo stage IV patients (17.2% v 7.3%; P < .001). CONCLUSION: We developed an IE system that accurately infers metastatic sites in multiple primary cancers from radiology reports. It has explainable methods and performs better than some clinical LLMs. The inferred metastatic phenotypes could enhance cancer research databases and clinical trial matching, and identify potential patients for oligometastatic interventions.

Comparison of an amplicon-based large panel next generation sequencing (NGS) assay with conventional testing methods for MET and HER2 amplification in lung and breast cancers.

The frequency of MET and HER2 amplification being detected by next generation sequencing (NGS) is increasing due to NGS being increasingly adopted for molecular profiling of cancers. However, the accuracy of NGS in detecting these gene amplifications remains uncertain due to conflicting reports in the scientific literature. We studied the accuracy of an amplicon-based large panel NGS assay in detecting MET and HER2 amplification in lung and breast cancers, respectively, by comparing it against conventional testing methods. Amongst 48 lung cancers, four of five cancers that were MET amplified on fluorescence in situ hybridisation (FISH) were classified as amplified on NGS while 42 of the remaining 43 non-amplified cancers were classified as non-amplified on NGS, giving a sensitivity of 80%, specificity of 97.7% and overall concordance of 95.8%. Of the 46 breast cancers tested, only six of the nine cancers that were HER2-positive on immunohistochemistry (IHC)/FISH were HER2-positive on NGS, while all the remaining HER2-negative cases were negative on NGS, giving a sensitivity of 66.7%, specificity of 100% and overall concordance of 93.5%. All the false-negative cases had low level gene amplification (MET:CEP7 or HER2:CEP17 FISH ratio of <3). The low sensitivity for HER2 amplification may be confounded by the small sample size and disproportionate number of cases with low level amplification. In summary, the NGS assay has good concordance with conventional testing methods but may be less sensitive in detecting low level gene amplification.

Combinatorial Hypofractionated Radiotherapy and Pembrolizumab in Anaplastic Thyroid Cancer.

Objectives Anaplastic thyroid cancer (ATC) is an aggressive disease associated with poor outcomes and resistance to therapies. Our study aim was to evaluate the activity of a combinatorial regimen of sandwich sequencing of pembrolizumab immunotherapy and hypofractionated radiotherapy (RT). Methods In this case series, patients with ATC received hypofractionated RT (QUAD-shot) and intravenous pembrolizumab 200mg every 3-4 weeks. Pembrolizumab was continued until disease progression or up till 24 months. Concurrent Lenvatinib treatment was allowed. Primary endpoint was best overall response (BOR) and progression-free survival (PFS). Additionally, we performed immune profiling of circulating T cells in a responder to investigate the immune response to our combinatorial treatment. Results At median follow-up of 32.6 months (IQR: 26.4-38.8), of a cohort of 5 patients, BOR was 80%; with 2 complete responses (CR) and 2 partial responses (PR). Patients who achieved CR remained disease-free at last follow-up. Median PFS was 7.6 months (IQR: 6.2-NR), and 1-year PFS and overall survival rate was 40% (95% CI: 13.7-100) for both. Treatment was well-tolerated, with mostly grade 1-2 adverse events. Immune profiling of one partial responder revealed an increase in activated CD4 and CD8 T cells post-QUAD-shot RT, which was further enhanced during the maintenance phase of pembrolizumab. Conclusions Herein, we reported a case series of 5 patients with ATC, with 2 long-term survivors who were treated with surgical debulking followed by QUAD-shot RT and pembrolizumab, possibly due to synergy of local and systemic treatments in activating anti-tumour immunogenic cytotoxicity. This regimen warrants further investigation in a larger cohort of patients.

Bioengineered hydrogels enhance ex vivo preservation of patient-derived tumor explants for drug evaluation.

Ex vivo patient-derived tumor slices (PDTS) are currently limited by short-term viability in culture. Here, we show how bioengineered hydrogels enable the identification of key matrix parameters that significantly enhance PDTS viability compared to conventional culture systems. As demonstrated using single-cell RNA sequencing and high-dimensional flow cytometry, hydrogel-embedded PDTS tightly preserved cancer, cancer-associated fibroblast, and various immune cell populations and subpopulations in the corresponding original tumor. Cell-cell communication networks within the tumor microenvironment, including immune checkpoint ligand-receptor interactions, were also maintained. Remarkably, our results from a co-clinical trial suggest hydrogel-embedded PDTS may predict sensitivity to immune checkpoint inhibitors (ICIs) in head and neck cancer patients. Further, we show how these longer term-cultured tumor explants uniquely enable the sampling and detection of temporal evolution in molecular readouts when treated with ICIs. By preserving the compositional heterogeneity and complexity of patient tumors, hydrogel-embedded PDTS provide a valuable tool to facilitate experiments targeting the tumor microenvironment.

Outcomes and prognostic factors of post-irradiation and de novo sarcomas of the head and neck: a histologically matched case-control study.

BACKGROUND: This study was designed to compare post-irradiation sarcomas (PIS) and de novo sarcomas (DN) of the head and neck in terms of tumor characteristics, prognostic factors, and survival outcomes. METHODS: All (N=83) head and neck sarcoma patients treated at National Cancer Centre, Singapore (Feb 2002-May 2011) were included: DN (N=60; 72%); PIS (N=23; 28%). Clinicopathologic features and outcomes of all patients and histologically matched pairs were compared. Prognostic factors were identified using univariate and multivariate analyses. RESULTS: Median age, gender, smoking status, and tumor size were not significantly different. Significant differences were seen in histology (most prevalent: PIS-sarcoma-NOS; DN-angiosarcoma) and tumor subsite (most prevalent: PIS-nasal cavity and sinuses; DN-skin). Median latency of PIS development was 16.7 years. PIS patients had shorter overall survival (OS) and disease-specific survival (DSS) compared with DN patients, most clearly seen on histologically matched pair analysis: 2-year OS (PIS: 54%; DN: 83%; P=0.028). Multivariate analyses showed that age>50 years (hazard ratio (HR)=3.68; P=0.007), ever-smokers (HR=2.79; P=0.017), and larger tumor-size (cm) (HR=1.12; P=0.045) were associated with worse OS, and age at >50 years (HR=2.77; P=0.04) and ever-smokers (HR=2.94; P=0.021) were associated with worse DSS. When treated with curative intent, no significant survival difference was noted between DN and PIS patients. CONCLUSIONS: In our cohort, PIS constituted 28% of head and neck sarcomas. Poorer prognosis traditionally associated with PIS compared with DN was not seen amongst patients treated with curative intent.

Prevalence and clinico-genomic characteristics of patients with TRK fusion cancer in China.

Neurotrophic tyrosine kinase (NTRK) fusions involving NTRK1, NTRK2, and NTRK3 were found in a broad range of solid tumors as driver gene variants. However, the prevalence of NTRK fusions in Chinese solid tumor patients is rarely reported. Based on the next-generation sequencing data from 10,194 Chinese solid tumor patients, we identified approximately 0.4% (40/10,194) of Chinese solid tumor patients with NTRK fusion. NTRK fusions were most frequently detected in soft tissue sarcoma (3.0%), especially in the fibrosarcoma subtype (12.7%). A total of 29 NTRK fusion patterns were identified, of which 11 were rarely reported. NTRK fusion mostly co-occurred with TP53 (38%), CDKN2A (23%), and ACVR2A (18%) and rarely with NTRK amplification (5.0%) and single nucleotide variants (2.5%). DNA-based NTRK fusion sequencing exhibited a higher detection rate than pan-TRK immunohistochemistry (100% vs. 87.5%). Two patients with NTRK fusions showed clinical responses to larotrectinib, supporting the effective response of NTRK fusion patients to TRK inhibitors.

Real-world clinical practice and outcomes in treating stage III non-small cell lung cancer: KINDLE-Asia subset.

Introduction: Stage III non-small cell lung cancer (NSCLC) is a heterogeneous disease requiring multimodal treatment approaches. KINDLE-Asia, as part of a real world global study, evaluated treatment patterns and associated survival outcomes in stage III NSCLC in Asia. Methods: Retrospective data from 57 centers in patients with stage III NSCLC diagnosed between January 2013 and December 2017 were analyzed. Median progression free survival (mPFS) and median overall survival (mOS) estimates with two sided 95% confidence interval (CI) were determined by applying the Kaplan-Meier survival analysis. Results: Of the total 1874 patients (median age: 63.0 years [24 to 92]) enrolled in the Asia subset, 74.8% were men, 54.7% had stage IIIA disease, 55.7% had adenocarcinoma, 34.3% had epidermal growth factor receptor mutations (EGFRm) and 50.3% had programmed death-ligand 1 (PD-L1) expression (i.e. PD-L1 ≥1%). Of the 31 treatment approaches as initial therapy, concurrent chemoradiotherapy (CRT) was the most frequent (29.3%), followed by chemotherapy (14.8%), sequential CRT (9.5%), and radiotherapy (8.5%). Targeted therapy alone was used in 81 patients of the overall population. For the Asia cohort, the mPFS and mOS were 12.8 months (95% CI, 12.2-13.7) and 42.3 months (95% CI, 38.1-46.8), respectively. Stage IIIA disease, Eastern Cooperative Oncology Group ≤1, age ≤65 years, adenocarcinoma histology and surgery/concurrent CRT as initial therapy correlated with better mOS (p < 0.05). Conclusions: The results demonstrate diverse treatment patterns and survival outcomes in the Asian region. The high prevalence of EGFRm and PD-L1 expression in stage III NSCLC in Asia suggests the need for expanding access to molecular testing for guiding treatment strategies with tyrosine kinase inhibitors and immunotherapies in this region.

Molecular testing in non-small cell lung cancer: A consensus recommendation.

Introduction: Lung cancer remains an important cause of cancer-related mortality in Singapore, with a greater proportion of non-smokers diagnosed with non-small cell lung cancer (NSCLC) in the past 2 decades. The higher prevalence of targetable genomic alterations in lung cancer diagnosed in Singapore compared with countries in the West, as well as the expanding therapeutic landscape for NSCLC in the era of precision medicine, are both factors that underscore the importance of efficient and effective molecular profiling. Method: This article provides consensus recommendations for biomarker testing for early-stage to advanced NSCLC. These recommendations are made from a multidisciplinary group of lung cancer experts in Singapore with the aim of improving patient care and long-term outcomes. Results: The recommendations address the considerations in both the advanced and early-stage settings, and take into account challenges in the implementation of biomarker testing as well as the limitations of available data. Biomarker testing for both tumour tissue and liquid biopsy are discussed. Conclusion: This consensus statement discusses the approaches and challenges of integrating molecular testing into clinical practice for patients with early- to late-stage NSCLC, and provides practical recommendations for biomarker testing for NSCLC patients in Singapore.

Safety, Tolerability, and Antitumor Activity of Zipalertinib Among Patients With Non-Small-Cell Lung Cancer Harboring Epidermal Growth Factor Receptor Exon 20 Insertions.

PURPOSE: Although several agents targeting epidermal growth factor receptor (EGFR) exon 20 insertions (ex20ins) have recently been approved by the US Food and Drug Administration, toxicities related to the inhibition of wild-type (WT) EGFR are common with these agents and affect overall tolerability. Zipalertinib (CLN-081, TAS6417) is an oral EGFR tyrosine kinase inhibitor (TKI) with a novel pyrrolopyrimidine scaffold leading to enhanced selectivity for EGFR ex20ins-mutant versus WT EGFR with potent inhibition of cell growth in EGFR ex20ins-positive cell lines. METHODS: This phase 1/2a study of zipalertinib enrolled patients with recurrent or metastatic EGFR ex20ins-mutant non-small-cell lung cancer (NSCLC) previously treated with platinum-based chemotherapy. RESULTS: Seventy-three patients were treated with zipalertinib at dose levels including 30, 45, 65, 100, and 150 mg orally twice a day. Patients were predominantly female (56%), had a median age of 64 years, and were heavily pretreated (median previous systemic therapies 2, range 1-9). Thirty six percent of patients had received previous non-ex20ins EGFR TKIs and 3/73 (4.1%) patients received previous EGFR ex20ins TKIs. The most frequently reported treatment-related adverse events of any grade included rash (80%), paronychia (32%), diarrhea (30%), and fatigue (21%). No cases of grade 3 or higher drug-related rash or diarrhea were observed at 100 mg twice a day or below. Objective responses occurred across all zipalertinib dose levels tested, with confirmed partial response (PR) observed in 28/73 (38.4%) response-evaluable patients. Confirmed PRs were seen in 16/39 (41%) response-evaluable patients at the dose of 100 mg twice a day. CONCLUSION: Zipalertinib has encouraging preliminary antitumor activity in heavily pretreated patients with EGFR ex20ins-mutant NSCLC, with an acceptable safety profile, including low frequency of high-grade diarrhea and rash.

Biomarker-directed, pembrolizumab-based combination therapy in non-small cell lung cancer: phase 2 KEYNOTE-495/KeyImPaCT trial interim results.

Although pembrolizumab confers clinical benefit in non-small cell lung cancer (NSCLC), only a subset of patients will respond due to a heterogenous tumor microenvironment. KEYNOTE-495/KeyImPaCT is an ongoing biomarker-directed, adaptively randomized phase 2 study investigating first-line pembrolizumab (200 mg every 3 weeks) + lenvatinib (20 mg daily), anti-CTLA-4 quavonlimab (25 mg every 6 weeks) or anti-LAG-3 favezelimab (200 mg or 800 mg every 3 weeks) in advanced NSCLC. Patients were categorized by T-cell-inflamed gene expression profile (TcellinfGEP) and tumor mutational burden (TMB) status and randomly assigned 1:1:1 to receive pembrolizumab + lenvatinib, pembrolizumab + quavonlimab or pembrolizumab + favezelimab. The primary outcome was investigator-assessed objective response rate (ORR) per Response Evaluation Criteria in Solid Tumors version 1.1 using pre-specified efficacy thresholds for each biomarker-defined subgroup (>5% (TcellinfGEPlowTMBnon-high (group I)), >20% (TcellinfGEPlowTMBhigh (group II) and TcellinfGEPnon-lowTMBnon-high (group III)) and >45% (TcellinfGEPnon-lowTMBhigh (group IV))). Secondary outcomes were progression-free survival, overall survival and safety. At data cutoff, ORR ranges were 0-12.0% in group I, 27.3-33.3% in group II, 13.6-40.9% in group III and 50.0-60.0% in group IV. ORR with pembrolizumab + lenvatinib in group III met the pre-specified efficacy threshold. The safety profile of each treatment arm was consistent with the known safety profile of each combination. These data demonstrate the feasibility of prospective TcellinfGEP and TMB assessment to study the clinical activity of first-line pembrolizumab-based combination therapies in advanced NSCLC. ClinicalTrials.gov registration: NCT03516981 .

Clinical efficacy and biomarker analysis of dual PD-1/CTLA-4 blockade in recurrent/metastatic EBV-associated nasopharyngeal carcinoma.

Single-agent checkpoint inhibitor (CPI) activity in Epstein-Barr Virus (EBV) related nasopharyngeal carcinoma (NPC) is limited. Dual CPI shows increased activity in solid cancers. In this single-arm phase II trial (NCT03097939), 40 patients with recurrent/metastatic EBV-positive NPC who failed prior chemotherapy receive nivolumab 3 mg/kg every 2 weeks and ipilimumab 1 mg/kg every 6 weeks. Primary outcome of best overall response rate (BOR) and secondary outcomes (progression-free survival [PFS], clinical benefit rate, adverse events, duration of response, time to progression, overall survival [OS]) are reported. The BOR is 38% with median PFS and OS of 5.3 and 19.5 months, respectively. This regimen is well-tolerated and treatment-related adverse events requiring discontinuation are low. Biomarker analysis shows no correlation of outcomes to PD-L1 expression or tumor mutation burden. While the BOR does not meet pre-planned estimates, patients with low plasma EBV-DNA titre (<7800 IU/ml) trend to better response and PFS. Deep immunophenotyping of pre- and on-treatment tumor biopsies demonstrate early activation of the adaptive immune response, with T-cell cytotoxicity seen in responders prior to any clinically evident response. Immune-subpopulation profiling also identifies specific PD-1 and CTLA-4 expressing CD8 subpopulations that predict for response to combined immune checkpoint blockade in NPC.

Real world efficacy of osimertinib in second line/beyond in patients with metastatic EGFR+ non-small cell lung cancer and role of paired tumour-plasma T790M testing at tyrosine kinase inhibitor resistance.

Background: Osimertinib is a third-generation epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) approved for use in EGFR-mutant lung cancer. We examined its performance in the second/subsequent line after resistance to first- and second-generation (1/2G) EGFR-TKI. Methods: We reviewed electronic records of 202 patients who received osimertinib from July 2015 to January 2019 in the second/subsequent line after progression on prior EGFR-TKI. Of these, complete data from 193 patients were available. Clinical data including patient characteristics, primary EGFR mutation, T790M mutation status, presence of baseline brain metastases (BM), first-line EGFR-TKI use, and survival outcomes were extracted, and results retrospectively analyzed. Results: Of 193 evaluable patients, 151 (78.2%) were T790M+ (T790M positive) with 96 (49.2%) tissue confirmed; 52% of patients received osimertinib in the second line setting. After median follow up of 37 months, median progression-free survival (PFS) of the entire cohort was 10.3 [95% confidence interval (CI): 8.64-11.50] months and median overall survival (OS) was 20 (95% CI: 15.61-23.13) months. Overall response rate (ORR) to osimertinib was 43% (95% CI: 35.9-50.3%); 48.3% in T790M+ vs. 20% in T790M- (T790M negative) patients. OS in T790M+ patients was 22.6 vs. 7.9 months in T790M- patients (HR 0.43, P=0.001), and PFS was 11.2 vs. 3.1 months respectively (HR 0.52, P=0.01). Tumour T790M+ was significantly associated with longer PFS (P=0.007) and OS (P=0.01) compared to tumour T790M- patients, however this association was not seen with plasma T790M+. Of the 22 patients with paired tumor/plasma T790M testing, response rate (RR) to osimertinib was 30% for those plasma T790M+/tumour T790M-, compared to 63% and 67% for those who were plasma T790M+/tumour T790M+ and plasma T790M-/tumour T790M+, respectively. By multivariable analysis (MVA), Eastern Cooperative Oncology Group (ECOG) performance status ≥2 was associated with shorter OS (HR 2.53, P<0.001) and PFS (HR 2.10, P<0.001), whereas presence of T790M+ was associated with longer OS (HR 0.50, P=0.008) and PFS (HR 0.57, P=0.027). Conclusions: This cohort demonstrated the efficacy of osimertinib in second line/beyond for EGFR+ (EGFR mutation-positive) non-small cell lung cancer (NSCLC). Tissue T790M result appeared more predictive of osimertinib efficacy compared to plasma, highlighting potential T790M heterogeneity and the advantage with paired tumor-plasma T790M testing at TKI resistance. T790M- disease at resistance remains an unmet treatment need.

Inferring cancer disease response from radiology reports using large language models with data augmentation and prompting.

OBJECTIVE: To assess large language models on their ability to accurately infer cancer disease response from free-text radiology reports. MATERIALS AND METHODS: We assembled 10 602 computed tomography reports from cancer patients seen at a single institution. All reports were classified into: no evidence of disease, partial response, stable disease, or progressive disease. We applied transformer models, a bidirectional long short-term memory model, a convolutional neural network model, and conventional machine learning methods to this task. Data augmentation using sentence permutation with consistency loss as well as prompt-based fine-tuning were used on the best-performing models. Models were validated on a hold-out test set and an external validation set based on Response Evaluation Criteria in Solid Tumors (RECIST) classifications. RESULTS: The best-performing model was the GatorTron transformer which achieved an accuracy of 0.8916 on the test set and 0.8919 on the RECIST validation set. Data augmentation further improved the accuracy to 0.8976. Prompt-based fine-tuning did not further improve accuracy but was able to reduce the number of training reports to 500 while still achieving good performance. DISCUSSION: These models could be used by researchers to derive progression-free survival in large datasets. It may also serve as a decision support tool by providing clinicians an automated second opinion of disease response. CONCLUSIONS: Large clinical language models demonstrate potential to infer cancer disease response from radiology reports at scale. Data augmentation techniques are useful to further improve performance. Prompt-based fine-tuning can significantly reduce the size of the training dataset.