Virtual Library
Start Your Search
Aaron C Tan
Author of
-
+
FP14 - Targeted Therapy - Clinically Focused (ID 252)
- Event: WCLC 2020
- Type: Posters (Featured)
- Track: Targeted Therapy - Clinically Focused
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
FP14.13 - Molecular Characterisation and Clinical Outcomes in RET Rearranged Non-Small Cell Lung Cancer (NSCLC) (ID 1463)
00:00 - 00:00 | Presenting Author(s): Aaron C Tan
- Abstract
Introduction
RET rearrangements are emerging as a targetable oncogenic fusion driver in 1-2% of NSCLC patients. Promising efficacy of novel selective RET tyrosine kinase inhibitors (TKI) such as selpercatinib and praseltinib have been demonstrated in early phase trials. However, the natural history of disease and the activity of different classes of systemic therapy remains to be defined. Furthermore, routine molecular testing for RET is not yet standard of care and the optimal method of testing is unclear. We present a comparative analysis of molecular profiling with FISH or NGS and correlations with clinical and treatment outcomes.
Methods
Patients diagnosed/treated with RET rearranged NSCLC at the National Cancer Centre Singapore between Apr 2014 and Mar 2020 were included. Multi-region whole exome sequencing (WES) was performed on one early stage pt. Baseline demographics and treatment outcomes were collected (median follow-up 20.3 months).
Results
A total of 64 patients were included, with median age 62 years (range 25-85), 56% were female, 77% Chinese ethnicity, 95% adenocarcinoma and 69% never smokers. RET rearrangement was detected by FISH in 30/34 (88%) patients, NGS in 40/43 (93%) patients, and with discordant results in 7/13 (54%) patients tested with both methods. Median TMB (n=17) was 5.4 mutations/Mb – TMB was high (>10) in 2 patients that were RET FISH positive but NGS negative. Known fusion partners were KIF5B-RET (62.5%), CCDC6-RET (30%), CNTNAP2-RET (2.5%), KIF5B-RET+CCDC6-RET (2.5%) and KIF5B-RET+THOC2-RET (2.5%). PD-L1 TPS was 0% in 6%, 1-49% in 23%, ≥50% in 18% and unknown in 52%. Multi-region WES (sectors=4) in an early stage patient revealed low TMB (median 1.6) and high intra-tumoral heterogeneity (pITH 0.66). EGFR co-mutation was detected in 8 (13%) patients, however in NGS RET positive patients it consisted only of uncommon EGFR mutations. Of 61 stage IIIB/IV or recurrent patients, prevalence of CNS metastases was 31%, and 92% received palliative systemic therapy – including platinum-pemetrexed chemotherapy (62%), immunotherapy +/- chemotherapy (28%), and multikinase (23%) or selective (57%) RET TKI therapy. Median PFS and ORR on chemotherapy was 7.7 months and 54%, on immunotherapy was 3.7 months and 29%, and on multikinase RET TKI was 3.3 months and 15% respectively. Response to immunotherapy was seen only in combination with chemotherapy. OS was prolonged in selective RET TKI treated versus untreated patients (median 49.3 vs 15.3 months; HR 0.16, 95%CI 0.06-0.40, p<0.0001), however was no different in immunotherapy treated versus untreated patients (median 37.7 vs 49.3 months; HR 1.30, 95%CI 0.53-3.19, p=0.53). OS was also prolonged in CCDC6-RET fusion versus KIF5B-RET fusion positive patients (median 113.5 vs 37.7 months; HR 0.12, 95%CI 0.04-0.38, p=0.009).
Conclusion
In RET rearranged NSCLC, selective RET TKI therapy is associated with improved survival outcomes, especially in CCDC6-RET positive patients. Immunotherapy has poor efficacy, associated with low TMB and PD-L1 expression. NGS and FISH testing methods also result in significant discordance.
-
+
MA04 - Health Policy and the Real World (ID 217)
- Event: WCLC 2020
- Type: Mini Oral
- Track: Health Services Research/Health Economics
- Presentations: 1
- Moderators:
- Coordinates: 1/29/2021, 16:45 - 17:45, Scientific Program Auditorium
-
+
MA04.06 - Clinical Characteristics and Outcomes in Advanced KRAS Mutant NSCLC – A Multi-Centre Collaboration in Asia (ATORG-005) (ID 3475)
16:45 - 17:45 | Author(s): Aaron C Tan
- Abstract
- Presentation
Introduction
KRAS driver mutations in advanced NSCLC have long been considered to be undruggable. However, promising efficacy data from early phase trials of novel therapies targeting KRAS have renewed focus on KRAS as an oncogenic driver. There is limited data on the prognostic and predictive significance of KRAS mutation subtypes. We present an interim analysis of a real world observational multi-centre study of advanced KRAS mutant NSCLC patients from five countries in Asia, conducted by the Asian Thoracic Oncology Research Group (ATORG).
Methods
Patients with advanced KRAS mutant NSCLC treated with at least one line of systemic therapy at tertiary centres in five Asian countries (China, India, Japan, Singapore, South Korea) between Jan 2014 and Dec 2018 were included. Baseline clinical characteristics, molecular profile and treatment outcomes were collected (median follow-up 35.5 months, 95%CI 28.7-50.3).
Results
A total of 155 patients were included in this interim analysis, with median age at advanced stage diagnosis 63 years (interquartile range [IQR] 56-70), 93% were ECOG 0-1, 70% were male and 64% were current or ex-smokers. In terms of ethnicity, 39% were Korean, 36% were Chinese, 15% were Japanese, 8% were Indian and 2% were Malay. Baseline histology was adenocarcinoma in 90%, squamous cell carcinoma in 4% and other histologies in 6%. KRAS mutation was detected by NGS in 141 (91%) patients, Sanger sequencing in 12 (8%) patients and RT-PCR in 2 (1%) patients. KRAS G12C (26%) was most common, followed by G12D (23%) and G12V (21%). The incidence of KRAS G12C mutation in patients with a smoking history was 35/99 (35%) compared with 6/56 (11%) in patients without any smoking history. Co-alterations were found with EGFR mutations (14%), ALK fusions (1%), ROS1 fusions (1%) and BRAF mutations (3%). PD-L1 TPS was 0% in 22%, 1-49% in 19%, ≥50% in 14% and unknown/not tested in 45%. Brain metastases were present at advanced stage diagnosis in 25% and lifetime prevalence was 35%. Patients received a median 2 lines of therapy. First-line systemic therapy consisted of chemotherapy alone (66%), targeted therapy (15%) or other therapies (19%). Median time to next treatment (TTNT) on first-line chemotherapy alone was 7.3 months (95%CI 5.0-9.5). Overall, the median TTNT for first-line and second-line therapy was 7.7 (95%CI 6.5–10.0) and 7.0 (95%CI 5.3–10.9) months, respectively. 63% of patients had died, and 37% of patients were still alive or lost to follow-up at the time of data cut-off. Median OS for the overall cohort was 21.6 months (95%CI 15.9-27.6). Median OS was greater in immunotherapy treated (alone or in combination at any line; 45%) versus non-immunotherapy treated (55%) patients (27.6 [95%CI 19.1-37.9] months versus 15.4 [95%CI 10.3-23.7] months, HR 1.8, 95%CI 1.2-2.7, logrank p=0.005).
Conclusion
In Asian KRAS mutant NSCLC, duration of first-line therapy and survival outcomes remain poor – emphasising the need for greater therapeutic options for patients with a KRAS driver mutation. Additional sites/countries are planned and recruitment to this study is ongoing.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
MA13 - Tumor Biology: Focus on EGFR Mutation, DNA Repair and Tumor Microenvironment (ID 214)
- Event: WCLC 2020
- Type: Mini Oral
- Track: Tumor Biology and Systems Biology - Basic and Translational Science
- Presentations: 1
- Moderators:
- Coordinates: 1/31/2021, 16:45 - 17:45, Scientific Program Auditorium
-
+
MA13.08 - Genomic and Transcriptomic Features of Distinct Resistance Trajectories in EGFR Mutant Non-Small Cell Lung Cancer (NSCLC) (ID 1461)
16:45 - 17:45 | Presenting Author(s): Aaron C Tan
- Abstract
Introduction
Despite the established role of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKI) in advanced EGFR mutant NSCLC, drug resistance inevitably ensues. More than 40% of resistant patients are EGFRT790M negative (T790M-), for which a diverse range of mechanisms have been described and there remains a paucity of treatment options. Here, we performed systematic multi-omics profiling and analysis of first- and second-generation (1G/2G) TKI resistant patients to better understand molecular features of EGFRT790M positive (T790M+) and T790M- disease.
Methods
We performed deep whole exome and transcriptome profiling on tumor tissue biopsies after resistance to 1/2G EGFR TKI. Tissue was obtained from 59 advanced EGFR mutant NSCLC pts treated at the National Cancer Centre Singapore, of which 38 (63%) were T790M+. Using paired tumor/normal whole exome data, we identified somatic mutations and copy number alterations with Mutect and GISTIC, respectively. Gene expression profiles were generated from paired-end tumor RNA-seq data using STAR and RSEM. Tumor heterogeneity, mutational signatures, and immune cell-type infiltration was inferred. Transcriptome profiles of cancer and stromal cells were inferred and analyzed using a novel transcriptome deconvolution approach, TUMERIC. Multiplex immunofluorescence and correlations with clinical outcomes was performed.
Results
Median age of the cohort was 59 years (range 41-79), 56% were female, 86% never smokers and 90% Chinese ethnicity. At baseline, 35 (59%) pts had EGFR exon 19 deletion, 22 (37%) pts had L858R mutation and 1 (2%) pt each had exon 19 insertion and L861Q mutation. Therapy prior to biopsy consisted of TKI alone (36%), TKI plus chemotherapy (12%) or sequential therapy with TKI and chemotherapy (52%). Median time to progression on 1/2G TKI was 10.3 (1.3-75.8) mths. Among previously reported resistance mechanisms, only MET alterations were significantly enriched in T790M- tumors (19% vs 3%, p=0.03). In contrast, we discovered genomic features associated with T790M status, including enrichment of TP53 alterations (86% vs 50%, p=0.01), 3q chromsomomal arm amplification (57% vs 13%, p<0.001), whole genome doubling (100% vs 82%, p=0.04) and proportional contribution of non-aging mutational signatures (p=0.003) in T790M- tumors. Strikingly, transcriptomic analysis revealed ubiquitous downregulation of adenocarcinoma lineage gene expression (NAPSA, NKX2-1/TTF-1, SFTA2, SFTA3) across cancer cells in T790M- tumors with corresponding upregulation of either squamous or neuroendocrine marker gene expression. Multiplex immunofluorescence confirmed decreased NAPSA and NKX2-1 expression in T790M- vs T790M+ tumors. Classifying tumors into validated transcriptomic subtypes (TRU/PP/PI) also revealed complete absence of the TRU subtype in T790M- tumors (0% vs 52%, p=0.006). Furthermore, tumors were clustered using GEP into immunehot (n=20) or immunecold (n=24) tumors, with distinct immune cell infiltrates and clinical outcomes according to T790M status. Finally, using a Bayesian statistical approach, we explored how T790M- and T790M+ disease could be predicted using comprehensive genomic and transcriptomic profiles of treatment-naïve patients.
Conclusion
We illustrate the interplay between genetic alterations, cell lineage plasticity and the tumor microenvironment – with distinct resistance trajectories associated with T790M status. In particular, histological transformation with lineage plasticity is underappreciated in T790M- tumors. This provides a framework for optimizing therapeutic strategies to circumvent resistance in EGFR mutant NSCLC.
-
+
OA01 - Established Drugs in Special Populations and New Drugs in Established Populations (ID 226)
- Event: WCLC 2020
- Type: Oral
- Track: Immunotherapy (Phase II/III Trials)
- Presentations: 1
- Moderators:
- Coordinates: 1/29/2021, 09:15 - 10:15, Scientific Program Auditorium
-
+
OA01.06 - Randomised Phase 2 Study of Nivolumab (N) Versus Nivolumab and Ipilimumab (NI) Combination in EGFR Mutant NSCLC (ID 1741)
09:15 - 10:15 | Author(s): Aaron C Tan
- Abstract
- Presentation
Introduction
Monotherapy PD1 inhibition is associated with low clinical efficacy in EGFR mutant NSCLC, in part due to preponderance for never smokers, low TMB and paucity of tumor T cell infiltration. Given the potential role for ipilimumab to enhance effector cells and deplete Treg, we sought to ascertain the efficacy, safety and pharmacodynamics of combination immune checkpoint inhibition in advanced EGFR mutant NSCLC.
Methods
In this open-label, phase 2 trial of nivolumab (N) monotherapy versus nivolumab-ipilimumab (NI) combination, patients with advanced EGFR mutant NSCLC who failed one line of standard EGFR TKI were randomized in a 1:1 ratio to nivolumab/ ipilimumab combination therapy or nivolumab monotherapy (with crossover permitted). Patients were stratified according to PDL1 status and presence of brain metastasis. The primary end point was overall response rate (ORR). Pretreatment and on treatment biopsies were acquired for exome and RNA sequencing, as were serial blood collections for immuno-monitoring using flow cytometry.
Results
Enrolment was terminated early due to futility. A total of 31 patients were enrolled: NI:16 (3 with paired biopsies), N:15 (4 crossover to NI, all with paired biopsies). 14 (45.2%) harboured EGFR T790M; 6 (19.3%) were smokers and 16 (51.6%) had a PDL1 expression level of 1% or more. 16 (51.6%) had CNS metastases at baseline. The 6-month PFS rate was 9.0% (2.5%, 31.8%) for the overall cohort, with no significant difference in survival between the 2 arms: median PFS was 1.31 months for N (1.22, NE) vs 1.22 months for NI (1.15, NE), p=0.96. Only 1 patient achieved PR in the entire cohort and received NI combination therapy. Of the 5 patients who derived clinical benefit (defined as best response of PR, or ongoing PR/SD at 6 months), 4 were T790M negative. Incidence of pneumonitis was 3.2% (n=1, G1). Other toxicities include IDDM (n=1), polymyositis (n=1), hypothyroidism (n=2), transaminitis (n=1) and rash (n=5). There were no G3-5 toxicities.
We next interrogated the immune landscape of EGFR TKI resistant NSCLC, and examined the impact of ipilumumab-nivolumab. 4/23 (17.4%) baseline samples were GEP high (2/4 were T790M positive), with enrichment for exhausted CD8, Treg and overexpression of IDO1. Of these 4 patients, 2 derived clinical benefit to N monotherapy. In sequential biopsies before and after exposure to checkpoint inhibitors, increase in infiltrating CD8 T cells was associated with clinical benefit.
Conclusion
While there were no significant toxicity concerns with combination checkpoint inhibition in TKI-resistant EGFR mutant NSCLC, durable response was not demonstrated in our study. High GEP alone did not seem to predict for response to checkpoint inhibitors in EGFR mutant NSCLC, possibly due an immunosuppressive tumor microenvironment.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
-
+
P37 - Pathology - Biomarker Testing (ID 107)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P37.18 - Lung NSCLC Molecular Diagnostic Comparison Between NGS and Multiplex PCR Assays (ID 3825)
00:00 - 00:00 | Author(s): Aaron C Tan
- Abstract
Introduction
Background: Lung NSCLC has led the targeted oncologic treatment landscape with several approval first line therapies and emerging targets. Single gene testing for diagnostic workup is becoming outpaced by panel and multigene testing. Next generation sequencing and multiplex PCR assays are able to provide practical solutions to the evolving clinical diagnostic requirements. In this study we compared the results and utility in clinical testing.
Methods
Methods: Sixty eight retrospective Lung NSCLC FFPE samples were used for both DNA and RNA assays by clinically validated and/or FDA approved NGS panels Thermofisher OncomineTM Focus assay / Oncomine comprehensive V3 Fusion ( 52 genes and 51 fusions drivers )and the AmoyDxR Pan Lung Cancer PCR Panel multiplex PCR assay ( 11 genes). Optimization of the PCR panel was done by the laboratory. Only gene targets present in both the NGS and PCR panel were used to show concordance. Other parameters for quality control including assay quality control and practical considerations were reviewed.
Results
Results: Preliminary analysis showed 93% (63/68)concordance for the DNA assay and 87% ( 59/68) concordance for the RNA assay. Up to 35% (21/68) of the samples showed suboptimal library quality control for the NGS assay but performed optimally on the PCR panel. There were differences in the requirements for input material, turn around time and costs.
Conclusion
Conclusion: There was high concordance rate of the results between the assays. RNA quality in FFPE may show some limitations and performed better on the PCR panel. Both assays have their own advantages. Further evaluation and consideration in the clinical treatment roles will depend on practical evaluation in the laboratory and in the broader clinical testing environment.
-
+
P38 - Pathology - Pathology/Staging (ID 108)
- Event: WCLC 2020
- Type: Posters
- Track: Pathology, Molecular Pathology and Diagnostic Biomarkers
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P38.03 - Immunohistochemical, Histologic and Genomic Characterisation of Early Stage Pulmonary Invasive Mucinous Adenocarcinoma (ID 1458)
00:00 - 00:00 | Presenting Author(s): Aaron C Tan
- Abstract
Introduction
Invasive mucinous adenocarcinoma (IMA) is an uncommon adenocarcinoma variant defined by goblet or columnar cells containing intracytoplasmic mucin and extensive extracellular mucin. Historically, IMAs have been associated with a poorer prognosis. However emerging evidence indicates a majority harbour potentially actionable driver oncogenes and IMA may be associated with downregulation of lung lineage-specifying transcription factor NKX2-1/TTF-1 resulting in a mid/hindgut phenotype. Morphologic features of IMA have also not been well characterised.
Methods
Patients (pts) with early stage resected pulmonary IMA (n=38) at our institution between 2009-19 were included. Morphologic features, including cell and stroma type, cytologic atypia, mitotic rate and spread through air spaces (STAS) were described. Immunohistochemistry for CDX2, CK7, CK20, Napsin-A and TTF-1 was conducted. Whole genome sequencing (WGS) was performed in a subset of pts. Histologic and genomic profiles were correlated.
Results
Median age was 66 years (range 42-82), 59% were male, 89% were of Chinese ethnicity and 55% were current/ex-smokers. Stage distribution after resection (AJCC 8th edition) was IA (50%), IB (16%), IIA (3%), IIB (18%), IIIA (8%), IIIB (3%) and IV (3%). Adjuvant therapy was given in 5 (13%) pts. Driver oncogenes were detected with mutations in EGFR (5%), and fusions in ALK (11%) and RET (3%). Histologically, most (61%) showed typical histomorphology, consisting of columnar tumour cells with basally located nuclei and apical mucin vacuole with mild degree of stratification/complexity, while the remainder (39%) showed a combination of typical mucinous subtype with other more cuboidal or polygonal cells and densely eosinophilic cytoplasm. The stroma was markedly fibrotic and desmoplastic in 68%, with a focal small fibrotic scar in 32%. The mitotic rate was low at 0-5 mitoses/10HPFs (92%), with rarely a higher mitotic rate corresponding to focal areas or more cytologic atypia (8%). Similarly, the cytologic atypia tended to be mild to moderate in 90%. Lymphovascular invasion (LVI) was rarely present (13%), and involvement of the pleura (26%) and parenchymal resection margin (5%) was uncommon. STAS was present in 68%. Pure mucinous tumours were associated with mainly columnar type tumour cells (r2=0.83, p<0.001), only mild-moderate cytologic atypia (r2=0.42, p=0.008) and absence of LVI (r2=0.48, p=0.002). Immunohistochemistry showed tumours positive for CK7 (100%), CK20 (84%), TTF-1 (71%), Napsin-A (63%) and CDX2 (45%). CDX2 positivity was associated with the presence of extensive fibrotic stroma (r2=0.38, p=0.02). After median 2.9 yrs follow-up, 6 (16%) pts had recurred – all stage II/III at baseline – with 2-year DFS of 84% and 5-year DFS of 76%. WGS in 9 (24%) pts revealed lung cancer oncogenic drivers in KRAS (33%), EGFR (11%), ERBB2 (11%), ERBB3 (11%) or no driver mutation/fusion detected (33%). Median TMB was 1.3 mutations/Mbp (range 0.03-2.1). Overall, the presence of known oncogenic drivers was associated with negative staining for CK20 (r2=0.45, p=0.005) and CDX2 (r2=0.31, p=0.06).
Conclusion
IMA is characterised by circumscribed nodules with extensive fibrotic stroma, mild to moderate cytological atypia and low mitotic rate, with good survival outcomes. CDX2 positivity is associated with fibrotic and desmoplastic stroma and absence of an oncogenic driver.
-
+
P59 - Tumor Biology and Systems Biology - Basic and Translational Science - Genomics (ID 197)
- Event: WCLC 2020
- Type: Posters
- Track: Tumor Biology and Systems Biology - Basic and Translational Science
- Presentations: 1
- Moderators:
- Coordinates: 1/28/2021, 00:00 - 00:00, ePoster Hall
-
+
P59.08 - THOR: Multi-Ethnic, Open Access Thoracic Cancer Genomics Resource (ID 2162)
00:00 - 00:00 | Author(s): Aaron C Tan
- Abstract
Introduction
While the multi-factorial process that give rise to lung cancer is driven by exogenous carcinogens and inherited genetic changes, there is growing evidence that racial and ethnic differences also influence incidence, mortality and treatment response. Existing cancer genomics resources do not provide ethnic variability in the genomics of lung cancer, owing to data availability and the prohibitive computational costs in preparing available genomics data for analysis. For example, one state-of-the-art platform, cBioportal, maintains a large repository of pan-cancer data along with tools to carry out omics analysis, but lacks uniformly processed Asian lung cancer samples to enable cross-cohort comparisons. Here, we introduce a flexible and user-friendly thoracic cancer genomics platform, THOR, that allows for user-friendly cross-cohort lung cancer multi-omics analyses, hosting thousands of uniformly processed Asian and Caucasian lung cancer samples from published cohorts.
Methods
Published and available Asian Lung Cancer multi-omics samples (N=567 samples) along with multi-omics samples from TCGA (N=1144), were downloaded and uniformly processed with bcbio-nextgen, using best-practices analysis for high-throughput DNA sequencing. All relevant and available clinical annotations were also downloaded. This comprised consensus variant calls from 5 algorithms, copy number variation analysis, alongside batch-corrected transcript quantification and estimation of immune-cell type composition. Genetic admixture and ethnicity was uniformly inferred across all patients. The resultant data was then uploaded and hosted on the Singapore Collaborative Oncology Data Portal (OncoSG, https://src.gisapps.org/OncoSG/).
Results
The uniform and best-practices processing of all data enables accurate and fair comparisons of multi-omics data across different cohorts and ethnicities. The OncoSG platform facilitates accessible and flexible web-based analytics, visualisation and collaborative sharing of the genomic, transcriptomic, and clinical data. Using THOR, the research community is now able embark on studying ethnic differences in lung cancer molecular profiles and patient outcomes.
Conclusion
THOR comprises standardised large-scale multi-omics data of multiple Asian and Caucasian lung cancer cohorts combined with advanced data analysis and visualisation tools, providing a powerful and accessible tool for studies of ethnic differences in lung cancer.