Search

BACKGROUND: Circulating tumour DNA (ctDNA) has been spotlighted as an attractive biomarker because of its easy accessibility and real-time representation of tumour genetic profile. However, the clinical utility of longitudinal ctDNA monitoring has not been clearly defined. METHODS: Serial blood samples were obtained from metastatic colorectal cancer patients undergoing first-line chemotherapy. ctDNA was sequenced using a targeted next-generation sequencing platform which included 106 genes. Changes in ctDNA profile and treatment outcome were comprehensively analysed. RESULTS: A total of 272 samples from 62 patients were analysed. In all, 90.3% of patients had detectable ctDNA mutation before treatment. ctDNA clearance after chemotherapy was associated with longer progression-free survival which was independent of radiological response (adjusted hazard ratio 0.22, 95% confidence interval 0.10–0.46). Longitudinal monitoring was able to detect ctDNA progression which preceded radiological progressive disease (PD) in 58.1% (median 3.3 months). Diverse resistant mutations (34.9%) and gene amplification (7.0%) at the time of PD were discovered. For 16.3% of the PD patients, the newly identified mutations could be potential candidates of targeted therapy or clinical trial. CONCLUSION: ctDNA profile provided a more accurate landscape of tumour and dynamic changes compared to radiological evaluation. Longitudinal ctDNA monitoring may improve personalised treatment decision-making.

Circulating tumor DNA (ctDNA) may reveal dynamic tumor status during therapy. We conducted serial ctDNA analysis to investigate potential association with clinical outcome in metastatic colorectal cancer (mCRC) patients receiving chemotherapy. Tissue KRAS/NRAS wild-type mCRC patients were enrolled and treated with first-line cetuximab-containing chemotherapy. ctDNA isolated from plasma were analyzed by next generation sequencing (NGS) with 16 targeted gene panel. Among 93 patients, 84 (90.3%) had at least 1 somatic mutation in baseline ctDNA samples (average 2.74). Five patients with KRAS or NRAS hotspot mutation in the ctDNA showed significantly worse progression-free survival (PFS) (p = 0.029). Changes in average variant allele frequency (VAF) in ctDNA showed significant correlation with tumor size change at the time of first response evaluation (p = 0.020) and progressive disease (PD) (p = 0.042). Patients whose average VAF decreased below cutoff (p p = 0.018). At the time of PD, 54 new mutations including KRAS and MAP2K1 emerged in ctDNA. ctDNA sequencing can provide mutation profile that could better reflect tumor mutation status and predict treatment outcome.

Background: Pancreatic cancer remains one of the most devastating malignancies due to lack of target therapeutic options for advanced disease. Liquid biopsy which analyzes tumor-related biomarkers in blood, has emerged as an innovative technology. The aim of this study was to identify circulating tumor DNA (ctDNA) mutations in pancreatic adenocarcinoma(PDAC) patients using multi-gene panels. Methods: Total of 45 patients (27 male, median age 63 years) blood samples were obtained at baseline then follow-up after 2 months samples were acquired in 16 patients at National Cancer Center, Korea from November 2021 to April 2022. CtDNA were analyzed by next-generation-sequencing (NGS) panel of 118 genes and the varient allele frequency (VAF) of detected genes were analyzed. Then the association between outcomes of 5FU treated response and survival (overall survival and progression-free survival; OS and PFS) and VAF reduction at follow-up were analyzed. Results: Total of 35 genes with 114 genetic alterations were detected and frequent mutations were as follows: KRAS (29.9%), TP53 (14.9%), GNAS(4.4%), SMAD4(4.4%), AR(3.5%), RNF434(3.5%), ATM(3.5%), BRCA2(2.6%) and PIK3CA(2.6%). The mean VAF level considering all pathogenic variants of 45 patients was 8.9%(range, 0 - 87.2) including 8 patients who were not detected ctDNA mutations in baseline, and the mean VAF was 4.2% (0 - 50.0) at follow-up in 16 patients. With association of clinical response, patients with partial response (PR) and stable disease (SD) were decreased VAF from 11% at baseline to 10% after 2 months, and three (60%) of patients showed clearance of ctDNA. Progressive disease (PD) patients were also decreased VAF 67% after 2 months, however, clearance of ctDNA was only observed in one patient (12.5%). Reduction of ctDNA more than 50% were associated with longer survival as PFS (> 50% reduction groups vs. no reduction group; 5 vs. 3 months, p value=0.903, OS (undefined vs. 6 months, p value =0.180). Conclusions: This study showed most common mutation was KRAS with several other mutations of ctDNA in PDAC, and outcomes represented association with ctDNA. Citation Format: SooBeen Heo, Sunhwa Park, Jung Won Chun, Yun-Hee Kim, Jun-Kyu Kang, Hwang-Phill Kim, Sang Myung Woo, Sun-Young Kong. Genetic alterations of circulating tumor DNA in advanced pancreatic cancer patients receiving 5-fluorouracil based chemotherapy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6683.

433 Background: Serial circulating tumor DNA (ctDNA) analysis enables us to detect minimal residual disease, to estimate recurrence and to trace evolution of cancer. However, the role of ctDNA in patients with advanced gastric cancer (AGC) who received palliative chemotherapy is still unclear. We performed this study to evaluate if serial ctDNA can predict clinical outcomes of chemotherapy and propose evolutionary findings during chemotherapy. Methods: The patients with histopathologically confirmed AGC and who received palliative chemotherapy were enrolled. Serial blood sample for ctDNA sequencing were obtained at baseline (BL) (before first-line chemotherapy), after first cycle and every 2-3 months until disease progression. ctDNA sequencing was performed for all samples using the next-generation sequencing (NGS) platform with a targeted gene panel including 106 genes. Single nucleotide variants (SNVs), insertions and deletions (Indels), and copy number variants (CNVs) were identified by ctDNA sequencing while serial ctDNA were monitored by tracing changes in the variants. ctDNA clearance was defined as no detection of any kind of variants or variant allele frequency (VAF) less than 0.1%. Results: A total of 50 patients were registered and 248 samples were collected. Median age was 62. HER2-positive GC were 5 (10.0%) and 49 received fluoropyrimidine plus platinum (1 did not receive chemotherapy). Response rate and median progression-free survival (PFS) were 42.9% and 10.7 months (95% CI, 7.9-13.6). Among 40 who were available for BL ctDNA, patients who were detected with ctDNA variant (D group) were 21 (52.5%). TP53 mutation was the most frequently observed genetic alteration in BL ctDNA (76.2%) and median copy numbers of ERBB2 in HER2-positive GC were 28 (range 2-42). D group were more frequently observed in tubular or papillary adenocarcinoma than in poorly cohesive carcinoma (69.0% vs. 9.1%, p = 0.001). Patients who were not detected with any kind of ctDNA variant (ND group) showed a trend in longer PFS, compared to D group, however it was not significant (10.5 vs 7.9 months, p = 0.29). In serial analysis during first-line chemotherapy, ctDNA clearance was observed in 14 (70.0%) out of 20 in D group. Patients with ctDNA clearance showed significantly longer PFS, compared with patients without ctDNA clearance (9.8 vs. 3.6 months, p < 0.001). Among 32 patients who had disease progression, increased VAF or re-development of previous BL ctDNA variants were observed in 16 (50.0%). Increased VAF or re-development of BL ctDNA were detected in 2.1 months earlier than clinical or radiologic deterioration. Conclusions: Serial ctDNA could predict clinical outcomes of chemotherapy and detect disease progression earlier than clinical or radiological findings in AGC. Further study for serial ctDNA is needed to find out the resistant mechanism and evolutionary process in cancer.

Introduction: The fragment size of cfDNA derived from tumor cells are known to be shorter than that from normal cells for not yet precisely known mechanisms. We hypothesized that the fragment size of cfDNA can be utilized to reliably estimate tumor burden in patients with mCRC undergoing palliative chemotherapy. Methods: cfDNA sequencing data from mCRC patients receiving first-line palliative chemotherapy at Seoul National University Hospital (Seoul, Korea) were used for this analysis. The patients were enrolled in a main study evaluating the dynamic changes of circulating tumor DNA mutational profiles during chemotherapy. Blood samples were obtained prior to chemotherapy and after every four cycles of chemotherapy until disease progression. cfDNA sequencing data from normal individuals were used as control data. cfDNA sequencing for both cancer patients and normal individuals were performed using AlphaLiquid® 100 target capture panel (IMBdx, Inc., Seoul, South Korea). AlphaLiquid® 100 is a tumor agnostic panel consist of 106 genes, including 10 gene fusion and MSI. Based on the panel sequencing data, the genetic alterations and the fragment size of cfDNA were calculated. The fragmentation ratio was defined by the ratio of the read fragment proportion in size range P1 (100 - 155 bp) and P2 (160 - 180 bp). Results: cfDNA sequencing data from 280 plasma samples from 62 mCRC patients and 50 healthy controls were used for analysis. Compared to cfDNA from healthy controls, the cfDNA fragment sizes from mCRC patients were significantly shorter (169.585 bp vs. 173.964 bp; p < 0.001). Comparing the fragment sizes by the mutational status, the fragment sizes of alleles with detected somatic mutations were significantly shorter than those with reference alleles (mean fragment size 155.853 bp in alleles with somatic mutations vs. 160.613 bp in reference alleles; p < 0.001), but such difference was not observed with germline mutations (mean fragment size 160.911 bp in alleles with germline mutations vs. 159.889 bp in reference alleles; p = 0.992). Further, the clonality inferred from the variant allele frequency (VAF) of somatic mutation was negatively correlated with the size of the cfDNA fragment. The read fragment proportion in size range P1 was significantly associated with the clonality (r = 0.86, p Conclusions: The fragmentation ratio calculated from cfDNA in mCRC patients can represent tumor burden in dynamic samples and potentially can be used as a biomarker. Citation Format: Jun-Kyu Kang, Hwang-Phill Kim, Yoojoo Lim, Su Yeon Kim, Tae-You Kim. cell-free DNA (cfDNA) fragment size as a potential quantitative biomarker for metastatic colorectal cancer (mCRC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5164.

8569 Background: The current standard of care (SOC) for limited disease (LD) small cell lung cancer (SCLC) is definitive concurrent chemoradiotherapy (CCRT). Despite the high response rate to SOC, up to 70% of the patients potentially experience disease recurrence and fail to show the prolonged clinical benefit. We investigated the feasibility of cell-free DNA (cfDNA) based genomic alteration and fragmentomic analysis using pre-and post-treatment samples to investigate the predictive value of disease recurrence in LD-SCLC. Methods: The blood sample from fifty LD-SCLC who were treated with definitive CCRT were collected before and after the treatment. Target sequencing, AlphaLiquid scan (IMBdx, Korea), composed of 106 target genes, was conducted using cfDNA and peripheral blood mononuclear cells. Based on the recurrence-free survival (RFS) interval of 12 months, patients were categorized into group with persistent response (PeR, n = 29) and non-PeR (n = 21). Fragmentomics analyses were collected using the proportion of cfDNA fragments sized in P1 (100 – 155 bp) and P2 (160 – 180 bp) ranges. Results: Initial analysis was conducted based on the gene of interest. Patients with RB1 mutation (n = 11) detected in follow-up sample demonstrated significant shorter RFS compared to the RB1 wild type (WT) patients (n = 39) (7.9 mon. vs. NR, P = 0.002). Among the patients who were available for our fragmentomics analysis in follow up samples (n = 19), non-PeR (n = 9) group were significantly high in P1 range ( P = 0.003) and low in P2 range ( P = 0.002) compared to PeR group (n = 10). AUCs using the fragment proportion in P1, proportion in P2, and the fragmentation ratio (proportion in P1 / proportion in P2), were 0.889, 0.911, and 0.889, respectively. The survival analysis using fragmentation ratio showed significantly longer RFS in fragmentation ratio low group compared to the high group (7.6 mon. vs. NR, P = 0.002). Integrating RB1 mutation status with fragmentation ratio, patients with both RB1 WT and fragmentation ratio low (n = 9) showed the most favorable outcomes ( P = 0.006). On the contrary, patients with either RB1 mutated or high in fragmentation ratio (n = 10) showed median RFS of rest than 12 months. Conclusions: In this study, we investigated the clinical feasibility of cfDNA detected mutation and size of the fragment in disease recurrence of LD-SCLC using the sample collected before and after definitive CCRT. It is observed that patients with either RB1 mutation or high fragmentation ratio detected from cfDNA after the completion of CCRT are likely to experience early disease recurrence. Our findings suggest that cfDNA could provide supplementary information in predicting early treatment failure and support the clinical decision in selecting high-risk patients who might need intense monitoring and additional consolidative treatment.

Analyzing cell-free DNA (cfDNA) as a source of circulating tumor DNA is useful for diagnosing or monitoring patients with cancer. However, the concordance between cfDNA within liquid biopsy and genomic DNA (gDNA) within tumor tissue biopsy is still under debate. To evaluate the concordance in a clinical setting, we enrolled 54 patients with metastatic colorectal cancer and analyzed their plasma cfDNA, gDNA from peripheral blood mononuclear cells (PBMC), and gDNA from available matched tumor tissues using ultra-deep sequencing targeting 10 genes (38-kb size) recurrently mutated in colorectal cancer. We first established a highly reliable cut-off value using reference material. The sensitivity of detecting KRAS hotspot mutations in plasma was calculated as 100%, according to digital droplet PCR. We could selectively detect clinically important somatic alterations with a variant allele frequency as low as 0.18%. We next compared somatic mutations of the 10 genes between cfDNA and genomic DNA from tumor tissues and observed an overall 93% concordance rate between the two types of samples. Additionally, the concordance rate of patients with the time interval between liquid biopsy and tumor tissue biopsy within 6 months and no prior exposure to chemotherapy was much higher than those without. The patients with KRAS mutant fragments in plasma had poor prognosis than those without the mutant fragments (33 months vs. 63 months; p

189 Background: ctDNA is an attractive alternative to tissue for its easy accessibility and real time representation of systemic tumor profile. We explored the utility of ctDNA profiling during the course of mCRC treatment. Methods: Serial blood samples were obtained from mCRC patients before and during first-line palliative chemotherapy at fixed intervals (after every four cycles) until confirmed disease progression. ctDNA was sequenced using targeted next-generation sequencing (NGS) platform with 106 genes. Changes of ctDNA profile and treatment outcome were comprehensively analyzed. Results: A total of 272 samples from 62 patients were analyzed. In the pre-treatment blood samples, 56 (90.3%) of patients had detectable ctDNA mutations including single nucleotide variants, short insertions/deletions and copy number changes (median 4.5 mutations/ patient, range 0 - 133). In 31 (50.0%) patients who had tissue NGS panel results performed in the clinic, overall concordance between mutations from ctDNA and tissue was 86.5%. In three patients, ctDNA mutational profiles were found to be completely different from tissue profiles. At further investigation, these patients were found to have a separate primary cancer in their colon. At the time of the first follow-up, most (98.0%) patients showed decrease of ctDNA from baseline, represented by average variant allele frequency (VAF) changes of all ctDNA mutations found. Clearance of ctDNA was achieved in 40 (78.4%) patients and was associated with longer progression-free survival (median PFS 11.8 moths in ctDNA clearance (+) vs. 4.7 months in ctDNA clearance (-), p < 0.001). The ctDNA clearance at the same time point was able to further discriminate the patients in same category by RECIST 1.1. Serial follow-up monitoring revealed three patterns of ctDNA changes at the time of clinical progressive disease (PD): 1) re-emergence or re-increase of baseline ctDNA mutations, 2) emergence of new resistance mutations, 3) radiologic PD without evidence of ctDNA progression. In the patients with detectable ctDNA at PD, the ctDNA progression preceded radiologic progression in 25 (58.1%) patients by median of 3.3 months. The patients in clinical PD without ctDNA progression showed different patterns of metastasis having mainly extrahepatic spread, while 77.8% of the patients with ctDNA progression had their progression confirmed in liver metastasis. Diverse resistant mutations and gene amplifications in PD patients were discovered by ctDNA sequencing. For seven (16.3%) of the PD patients, the newly identified mutations could be potential candidates of targeted therapy or clinical trial. Conclusions: ctDNA profile provided additional information to conventional evaluation methods and reflected dynamic changes. ctDNA monitoring may improve precise treatment decision-making for individual patients.

There is a close link between Fusobacterium nucleatum and colorectal cancer (CRC) tumorigenesis and chemoresistance. However, the genetic characteristics and clinical significance of CRC related with F. nucleatum remains unclear. This study evaluated the relationship between F. nucleatum, pathway mutation, and patient prognosis. Fusobacterium nucleatum amount in the tumor tissue and adjacent normal tissue were measured by quantitative polymerase chain reaction in adjuvant (N = 128) and metastatic (N = 118) cohorts. Patients were divided into binary (F. nucleatum-high and F. nucleatum-low) according to F. nucleatum amount. Targeted next-generation sequencing of 40 genes included in the 5 critical pathways (WNT, P53, RTK-RAS, PI3 K, and TGF-β) was performed in the adjuvant cohort. Patients with MSI-H and CIMP-H had higher amount of F. nucleatum in tumor tissue. Fusobacterium nucleatum-high patients had higher rates of transition mutation and C to T (G to A) nucleotide change regardless of MSI status. In addition, mutation rate of AMER1 and ATM genes, and TGF-β pathway was higher in F. nucleatum-high patients. Fusobacterium nucleatum-high was associated with poor overall survival (OS) in the palliative cohort (26.4 vs. 30.7 months, p = 0.042). Multivariate analysis revealed F. nucleatum-high as an independent negative prognostic factor for OS [adjusted hazard ratio of 1.69 (95% confidence interval 1.04–2.75), p = 0.034]. However, F. nucleatum amount was not associated with recurrence in the adjuvant cohort. F. nucleatum-high was associated with poor survival in metastatic CRC. In addition, we identified mutational characteristics of colorectal cancer according to F. nucleatum amount.

Epidermal growth factor receptor (EGFR) blockade is efficacious in KRAS wild-type of colorectal cancer, but acquired resistance is inevitably generated. Here, we established the 5 acquired resistant single cell models (CR cell lines) to cetuximab and investigated the mechanism of acquired resistance using single cell-based whole-exome sequencing and transcriptional profiling. Of these, 4 cell lines have KRAS Q61L mutation and the other has BRAF V600E mutation. From gene enrichment analysis (GSEA), we found that epithelial to mesenchymal transition (EMT) occurs and CXCL chemokine family, which are known as secreted proteins, are overexpressed in the CR cell lines compared with parental cells. Interestingly, we found that CXCL chemokines were strongly secreted/expressed in KRAS/BRAF mutant cell lines and plasma/serum samples from colorectal cancer patients who acquired KRAS/BRAF mutation during cetuximab therapy. CXCL chemokines are regulated coordinately by Slug and NF-B transcription factors and trans-activate EGFR through CXCR2/MMP1/sHB-EGF axis. This process activates MAPK pathway via autocrine. To overcome the resistance to cetuximab, combination treatment of cetuximab and MEK/BRAF inhibitor had a larger antitumor effect compared to the single agents, decreasing the expression of MAPK pathway and secretion of CXCL chemokines. This finding highlights that CXCL chemokines, as a secreted proteins circulating in the bloodstream, show correlation with KRAS/BRAF mutation and could be clinical biomarkers for predicting resistance to anti-EGFR therapy in CRC. Citation Format: Ye-Lim Park, Hwang-Phill Kim, Jun Kyu Kang, Yoo Joo Lim, Sang-Hyun Song, Sae-Won Han, Tae-You Kim. Single cell-driven secreted proteins as a surrogate biomarker for Cetuximab therapy in colorectal cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 419.

Purpose RNA editing generates protein diversity by altering RNA sequences in coding regions without changing the overall DNA sequence. Adenosine-to-inosine (A-to-I) RNA editing events have recently been reported in some types of cancer, but they are rare in human colorectal cancer (CRC). Therefore, this study was conducted to identify diverse RNA editing in CRC. Materials and methods We compared transcriptome data of 39 CRC samples and paired adjacent tissues from The Cancer Genome Atlas database to identify RNA editing patterns in CRC, focusing on canonical A-to-I RNA edits in coding sequence regions. We investigated nonsynonymous RNA editing patterns by comparing tumor and normal tissue transcriptome data. Results The number of RNA edits varied from 12 to 42 per sample. We also observed that hypoand hyper-RNA editing patterns were distinguishable within the samples. We found 10 recurrent nonsynonymous RNA editing candidates in nine genes (PDLIM, NEIL1, SRP9, GLI1, APMAP, IGFBP7, ZNF358, COPA, and ZNF587B) and validated some by Sanger sequencing and the inosine chemical erasing assay. We further showed that editing at these positions was performed by the adenosine deaminase acting on RNA 1 enzyme. Most of these genes are hypoedited in CRC, but editing of GLI1 was increased in cancer tissues compared with normal tissues. Conclusion Our results show that nonsynonymous RNA editing patterns can be used to identify CRC patients and could serve as novel biomarkers for CRC.

The cell-free DNA (cfDNA) represents a minimally invasive and alternative source of tumor DNA for molecular profiling. Despite next-generation sequencing (NGS) technique is qualified for genotyping cancer using cfDNA as a noninvasive method, it has caused problems as sequencing error and reproducibility. cfDNA in plasma and gDNA of Peripheral Blood Mononuclear Cells (PBMC) were isolated from each 54 advanced colorectal cancer patients. 39 available tumor tissues were isolated from same patients. Deep target-sequencing was performed with paired-end library enriched exons of 10 genes which are recurrently mutated in colorectal cancer. To reduce sequencing error, we devised ‘Denoising’ and calculated concordance of somatic variants between cfDNA and tumor tissue sequencing data. In addition, correlation of concordance data was analyzed with the clinical information. As a result, we selectively could detect clinically important somatic alteration among low/high variant allele frequency (0.31%~79.42%). For somatic alteration of 10 genes, sensitivity, specificity and accuracy were increased from 84.5%, 74.6% and 76.9% to 87.6%, 92.0% and 91.1% respectively after ‘Denoising’. On the other hand, patients with high cfDNA concentration(>50ng/ml) had higher somatic mutant fragments and larger metastatic lesion in liver than patients who have low cfDNA concentration. Our study showed that denoised deep target-sequencing is a suitable method for cfDNA genotyping and provides insights into strategies for monitoring metastatic lesion of advanced colorectal cancer. Citation Format: Jun-Kyu Kang, Hwang-Phill Kim, Seul-Ki Cheon, Ye-Lim Park, Yoojoo Lim, Sae-Won Han, Tae-You Kim. Noninvasive approach to assess metastatic lesion of advanced colorectal cancer by denoised deep target sequencing [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 1800. doi:10.1158/1538-7445.AM2017-1800

KRAS is a frequently mutated gene in colorectal cancer. In addition, PIK3CA mutations commonly co-exist with KRAS mutations and lead to additive activation of the PI3K/mTOR signaling pathway. Here, we investigated preclinical activity of gedatolisib, a PI3K/mTOR dual inhibitor, to identify mechanism of inhibition of PI3K/mTOR in 28 colorectal cancer (CRC) cells. Cells specifically with PIK3CA mutation were sensitive while those with KRAS mutation were resistant to gedatolisib. 9 out of 28 CRC cells harbor PIK3CA & KRAS co-mutation and 7 out them were shown to be sensitive to gedatolisib. However, HCT15 and LS174T cells were resistant. We identified that resistant cell lines have high activity of GSK3B and TCF7 frameshift mutation (465insertC466;H155fs*), which functions as positive regulator of WNT/b-catenin pathway. The effects of GSK3B-knockdown showed decreased activity of mTOR downstream molecules in gedatolisib-treated resistant cells. Interestingly, these effects also caused a decrease in activity of WNT/b-catenin pathway. In addition, combination treatment of gedatolisib and CHIR-99021, GSK3B inhibitor, resulted in significantly enhanced cytotoxicity against gedatolisib-resistant TCF7 frameshift mutant cells. Taken together, these show that aberrant regulation of WNT/b-catenin pathway and high activity of GSK3B by TCF7 frameshift mutation cause resistance to PI3K/mTOR dual inhibitor. Inhibition of GSK3B activity in colorectal cancer cells with KRAS and PIK3CA co-mutations increases sensitivity to PI3K/mTOR dual inhibitor. Citation Format: Ye-Lim Park, Hwang-Phill Kim, Seul-Ki Cheon, Jun-Kyu Kang, Yoojoo Lim, Sang-Hyun Song, Sae-Won Han, Tae-you Kim. Activation of WNT/b-catenin signaling results in resistance to PI3K/mTOR dual inhibitor in co-existing KRAS and PIK3CA mutant colorectal cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5214. doi:10.1158/1538-7445.AM2017-5214

Although MEK blockade has been highlighted as a promising anti-tumor drug, it has poor clinical efficacy in KRAS mutant colorectal cancer. Several feedback systems have been described in which inhibition of one intracellular pathway leads to activation of a parallel signaling pathway, thereby decreasing the effectiveness of single-MEK targeted therapies. In this study, we describe a feedback mechanism in which MEK inhibition leads to activation of macrophage migration inhibitory factor (MIF)-induced stat3 signaling pathway in KRAS mutant colorectal cancer (CRC) cells. We found that KRAS mutant CRC cells with refametinib, MEK inhibitor, induced MIF secretion and resulted in activation of Stat3. MIF knockdown by siRNA partially restored sensitivity to refametinib in KRAS mutant cells. In addition, combination with refametinib and 4IPP, a MIF inhibitor, effectively reduced the activity of stat3 and MAPK, more than single agent treatment. As a result, combined therapy was found to exhibit a synergistic growth inhibitory effect against refametinib-resistant cells by downregulating MIF expression. These results reveal that MIF-induced stat3 activation evoked an intrinsic resistance to refametinib. Our results provide the basis for a rational combination strategy against KRAS mutant colorectal cancers, predicated on the understanding of cross-talk between the MEK and MIF pathways. Citation Format: Seul-Ki Cheon, Hwang-Phill Kim, Ye-Lim Park, Si Hyun Lee, Jun-Kyu Kang, Yoojoo Lim, Sang-Hyun Song, Sae-Won Han, Tae-You Kim. MIF-induced stat3 activation promotes resistance to MEK blockade in KRAS mutant colorectal cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3167. doi:10.1158/1538-7445.AM2017-3167