Siewert III Adenocarcinoma: Still Searching for the Right Treatment Combination

It remains uncertain whether Siewert III tumors should be treated as esophageal or gastric cancers. Neoadjuvant therapy has been shown to improve survival in both esophageal and gastric trials. Randomized control trials comparing neoadjuvant chemotherapy versus chemoradiation should help define the most optimal treatment regimen. Surgical treatment follows general oncology principals: resect to negative margins with complete lymph node dissection, and, the extent of resection often extends more proximal onto the esophagus in addition to the total/subtotal gastrectomy.

Guidelines Insights: Hepatobiliary Cancers, Version 2.2019

The NCCN Guidelines for Hepatobiliary Cancers provide treatment recommendations for cancers of the liver, gallbladder, and bile ducts. The NCCN Hepatobiliary Cancers Panel meets at least annually to review comments from reviewers within their institutions, examine relevant new data from publications and abstracts, and reevaluate and update their recommendations. These NCCN Guidelines Insights summarize the panel's discussion and updated recommendations regarding systemic therapy for first-line and subsequent-line treatment of patients with hepatocellular carcinoma.

COVID-19 pandemic and impact on cancer clinical trials: An academic medical center perspective

The COVID-19 pandemic changed health-care operations around the world and has interrupted standard clinical practices as well as created clinical research challenges for cancer patients. Cancer patients are uniquely susceptible to COVID-19 infection and have some of the worst outcomes. Importantly, cancer therapeutics could potentially render cancer patients more susceptible to demise from COVID-19 yet the poor survival outcome of many cancer diagnoses outweighs this risk. In addition, the pandemic has resulted in risks to health-care workers and research staff driving important change in clinical research operations and procedures. Remote telephone and video visits, remote monitoring, electronic capture of signatures and data, and limiting sample collections have allowed the leadership in our institution to ensure the safety of our staff and patients while continuing critical clinical research operations. Here we discuss some of these unique challenges and our response to change that was necessary to continue cancer clinical research; and, the impacts the pandemic has caused including increases in efficiency for our cancer research office.

Molecular correlates of PD-L1 expression in patients (pts) with gastroesophageal (GE) cancers

4558

Background: The increased PD-L1 expression evaluated by combined positive score (CPS) is associated with improved efficacy of immunotherapy in GE cancers. The impact of tumor molecular alterations on PD-L1 expression is still not well-studied. We aimed to characterize specific molecular features of tumors with different CPS levels in GE cancers. Methods: 2,707 GE tumors [1,662 gastric/GE junction adenocarcinoma (GA), 856 esophageal adenocarcinoma (EA), 75 esophageal squamous (ES) and 114 GE unspecified] collected between 2000.8 and 2019.7 were analyzed using NextGen DNA sequencing (NGS), immunohistochemistry (IHC) and fragment analysis (FA) (Caris Life Sciences, Phoenix, AZ). Tumor mutation burden (TMB) was calculated based on somatic nonsynonymous missense mutations. dMMR/MSI status was evaluated by a combination of IHC, FA and NGS. PD-L1 expression measured by IHC (22c3) was evaluated by CPS scores. Molecular alterations were compared in three groups (CPS ≥ 10, H; CPS = 1~9, M; CPS = 0, L) using Fisher-Exact or Chi-square and adjusted for multiple comparison by Benjamini-Hochberg. Significance was determined by adjusted (adj) p < .05. Results: Overall, CPS-H, M, and L were seen in 18% (n = 494), 28% (n = 765) and 53% (n = 1,448) of GE tumors respectively. CPS-H was the most prevalent in ES (43%) followed by GA (19%) and lowest in EA (14%). Overall, TMB was similar between CPS-L and M, but was significantly increased in H (average TMB = 8.4 vs. 8.6 vs. 11 mt/MB, adj p < .0001); the effect was seen in EA and GA, but not in ES. An overall significant association between MSI/dMMR status and PD-L1 expression levels was seen (2%, 3.2% and 12% in CPS-L, M and H, adj p < .05) in GE tumors; the significance was seen in GA, but not in EA or ES. Amplifications of PD-L1 (H: 1.5%, M: 0.1% and L: 0) and PD-L2 (H: 1.1%, M: 0.1%, L: 0) were the highest in CPS-H, while ASPSCR1 (H: 0, M: 0, L: 1%) and TNFRSF14 (H: 0, M: 0.4, L: 2%) were the lowest (adj p < .01). Genes involved in epigenetic modification (top 5: ARID1A, ASXL1, BCL9, BCOR, CREBBP), MAPK ( KRAS, MAP2K1) and mismatch repair ( MLH1, MSH6) had the highest mutation rates in CPS-H, compared to M and L ( p < .0001). In contrast, CDH1 had higher mutation rates in CPS-L (12%), compared to M and H (5% and 5%) ( p < .0001). Conclusions: This is the largest study to investigate the distinct molecular landscape of pts with different PD-L1 expression levels in GE cancers. Our data may provide novel insights for pt selection (e.g. pts with gene mutations involved in epigenetic modification) and the development of rational combination immunotherapy (e.g. drugs targeting MAPK pathway).

Enrichment of alterations in targetable molecular pathways in KRAS wild-type (WT) pancreatic cancer (PC)

4629

Background: Genomic profiling has identified KRAS mutations in 88-90% of PC. KRAS WT tumors represent a molecularly heterogeneous group that may harbor targetable alterations (TA). We studied KRAS WT PC using NextGen sequencing (NGS) and whole transcriptome sequencing (WTS) to characterize the molecular landscape of this unique group and to assess the prevalence of TA. Methods: A total of 1164 PC tumors were tested at Caris Life Sciences by NGS (592 genes), WTS (NovaSeq), IHC and fragment analysis. Comparison of KRAS WT vs. mutant (MT) was done by Fisher-Exact or Chi2 and was corrected for multiple tests. Results: The KRAS WT cohort included 144 tumors (12.4%). No differences were seen in gender (female: 46% in both WT & MT) and age (median: 66 & 67) compared to KRAS MT. In KRAS WT tumors, targetable fusions tested by WTS and pathogenic mutations by NGS were seen in 22% (32 of 144) and 52% (75 of 144) respectively; potentially targetable amplifications (amp) were seen in 5 additional tumors. No TA were seen in 22% of WT tumors. Key alterations are in Table. Alterations inducing MAPK activation, including BRAF, RAF1, NF1 and GNAS changes were seen in 38 (26%) tumors. Frequent alterations were seen in FGFR genes (11 tumors), MET (4, including 1 exon 14 skip), and ERBB receptor and ligands (10). Fusions in ALK, ROS1, RET and NOTCH1 were seen (8), largely exclusive of other drivers. Wnt, PI3K, chromatin remodeling (CR) and DDR changes were common and sometimes seen concurrent with other alterations. Compared to KRAS MT, no difference of PD-L1 or TMB-H was seen. BRAF, APC, PBRM1, CTNNB1 mutations, MDM2 amp, gene fusions and MSI-H/dMMR were all more frequent in KRAS WT tumors (corrected p < 0.05). Conclusions: KRAS WT PC is enriched with TA (e.g., BRAF, ALK, ROS1, NRG1, MSI-H). The use of WTS in combination with NGS identifies activated molecular pathways in the majority of KRAS WT tumors. Based on our findings, comprehensive profiling of PC at the DNA and RNA level is recommended to provide patients with therapeutic opportunities beyond standard treatments. [Table: see text]

The landscape of POLE variants in colorectal and endometrial tumors: Correlation with microsatellite instability (MSI) and tumor mutation burden (TMB)

e13538

Background: Polymerase epsilon (POLE) is a major replicative DNA polymerase. Somatic POLE pathogenic variants (PV) are prevalent in endometrial cancer (EC) and in germline predispose to colorectal cancer (CRC), EC, and possibly other cancers (CA). PVs in the exonuclease domain (ExoD) [amino acid (AA) 268-471] lead to CAs with exceptionally high TMB. PV and uncertain variants (VUS) outside ExoD are sometimes concurrent with an ExoD PV and/or MSI. We hypothesized that the presence of non-ExoD variants may further increase POLE-associated mutation rate and tumor mutational burden. Methods: We retrospectively examined 1870 CRC and 4481 EC genomic profiles conducted by Caris Life Sciences (6/2016-6/2019). All patients had a 592-targeted gene somatic panel. Profiles with a POLE variant (PV or VUS) were analyzed. Median TMB (TMB, in mutations/megabase) was dichotomized to low/intermed ( < 17) vs high (>17). Tumors were grouped by: single ExoD PV, ExoD PV plus another variant (PV or VUS), or no ExoD PV. Known CRC/EC ExoD PV drivers were identified (Campbell et al, Cell 2017): D275G, P286R, S297F/Y, F367C/L/V, V411L, L424F, P436R/S/Y, M444K/L, A456P, S459F/Y, S461L/P, A465V. Kruskal-Wallis and chi-square tests were used. Results: Overall 4.5% CRC (80/1870) and 6.5% EC (303/4481) samples had POLE variants (Table). High TMB was seen in 56.3% CRC and 53.3% EC. In both CRC/ECs, TMB was higher in tumors with an ExoD PV and a 2nd variant compared to those with a solitary ExoD PV or no ExoD PV (both p < 0.001). MSI was more prevalent in CRC and EC with high TMB but no ExoD PV vs those with either high TMB and an ExoD PV, or low/intermed TMB and no ExoD PV (both p < 0.001). In both CRC/ECs, several ExoD PV associated with very high TMB when non-ExoD regions of POLE contained recurrent variant clusters: AA 1906 (TMB 225); AA 1826-7 (TMB 243); AA 1380-2 (TMB 229). Conclusions: In CRC/ECs, POLE ExoD PV and MSI appear to drive TMB in distinct and largely non-overlapping ways. Non-ExoD POLE variants may synergize with ExoD PVs to further increase mutation rates. [Table: see text]

The landscape of MAP3K1/MAP2K4 alterations in gastrointestinal (GI) malignancies

4113

Background: Inactivating alterations in MAP3K1/MAP2K4 occur in various solid tumors, sensitize cancer models to MEK inhibitors, and have co-mutation partners which may enable therapeutic targeting. Methods: We retrospectively reviewed 20290 GI malignancy patients (pts), comprised of 9986 colorectal carcinoma (CRC) and 10304 non-CRC, whose tumors were profiled with Caris Life Sciences from 2015-2019. Profiling included immunohistochemistry (IHC) with programmed death ligand-1 (PD-L1), next-generation sequencing (NGS), tumor mutational burden (TMB) and deficient mismatch repair or microsatellite instability-high status (dMMR/MSI-H). Results: MAP3K1/MAP2K4-alteration ( MAP3K1/MAP2K4-MT) was more frequent in CRC than non-CRC pts (2.0% v. 1.2%, p<0.0001), with truncating mutations representing the majority of lesions along both genes. While MAP3K1/MAP2K4-MT CRC pts were similar in age and gender to wild-type (WT), mutated non-CRC pts were older (median age 69 v. 65 years) and more likely female (51% v. 42%) compared to WT (both p<0.05). MAP3K1/MAP2K4-MT CRC (25% v. 7%) and non-CRC (30% v. 3%) were more frequently dMMR/MSI-H than WT pts (both p<0.0001). MAP3K1/MAP2K4-MT CRC cases were affiliated with higher TMB and similar rate of PD-L1 expression compared to WT. A higher rate of MAP3K1/MAP2K4-MT CRC pts were right-sided (36% v. 22%, p<0.0001) and transverse (8% v. 4%, p<0.05) compared to WT, whereas a higher frequency of WT cases were left-sided (20% v. 28%, p<0.05) and rectal (15% v 23%, p<0.05). Of microsatellite stable (MSS) CRC pts, those with MAP3K1/MAP2K4-MT were more likely PIK3CA (26% v. 17%) and APC (85% v. 78%) and less-likely TP53 (64% v. 77%) co-mutated versus WT MSS pts (all p<0.05); no difference was seen in BRAF V600E, ERBB2/ ERBB3 or KRAS co-mutation rate in MSS pts. In both all-comers and MSS CRC, MAP3K1/MAP2K4-MT pts were more frequently co-mutated than WT with ARID1A, POLE, ATM, BRCA2 and PIK3R1 (all ≥7% of MAP3K1/MAP2K4-MT pts, p<0.0001). A higher frequency of all-comer non-CRC GI malignancy pts with MAP3K1/MAP2K4-MT were co-mutated with PIK3CA (13% v. 6%), ERBB2/ERBB3 (8% v. 3%) or APC (13% v. 5%) compared to WT (all p<0.01). For MSS non-CRC GI cases, ARID1A (50% v. 30%) and SMAD4 (21% v. 12%) were more frequently co-mutated in MAP3K1/MAP2K4-MT versus WT pts (all p<0.05). Conclusions: Truncating MAP3K1/MAP2K4 alterations occur in nearly 2% of GI malignancy pts and are more commonly associated with dMMR/MSI-H than WT. Potentially targetable co-mutation partners implicated in MAPK and PI3K pathways as well as POLE, BRCA2 and ATM warrant further evaluation.

Comprehensive analysis of KRAS variants in patients (pts) with pancreatic cancer (PDAC): Clinical/molecular correlations and real-world outcomes across standard therapies

4641

Background: Approximately 90% of PDAC tumors are driven by activating KRAS mutations. The biological and clinical impact of common KRAS variants (e.g. G12D, G12V, G12R) and less common variants (e.g. G12C, Q61H, Q61R) remains largely unknown despite the emergence of variant-specific treatment strategies. Methods: We retrospectively analyzed real-world outcomes from 1475 PDAC pts who underwent molecular profiling via the Know Your Tumor program. Overall survival (OS) and progression-free survival (PFS) were analyzed by choice of 1st line standard therapies. Outcomes in pts with specific KRAS mutations were compared against the KRAS G12D cohort using Cox regression. Based on our prior data, tumor profiles with actionable molecular findings (DDR mutations or other drivers) were evaluated separately. Results: The prognostic/predictive value of specific KRAS variants revealed differences in real-world outcomes (Table). OS was greater in pts with KRAS G12V and G12R variants, as was PFS on 5FU-Based Therapy (e.g. FOLFIRINOX) but not for Gemcitabine/nab-Paclitaxel. Opposing trends were noted for KRAS Q61. Pts with KRAS wild type tumors as well as both actionable subgroups also had an improved OS. Conclusions: In this large national dataset, we demonstrate that KRAS mutation status and specific variants appear to be prognostic as well as predictive in pancreatic cancer. [Table: see text]

Somatic alterations of NF1 in colorectal cancer

4066

Background: NF1 encodes neurofibromin, which is a key GTPase-activating protein that downregulates RAS activation. Inactivating mutations in NF1 result in sustained activation of RAS signaling, a key driver for development of colorectal cancer (CRC), and have been suggested to be a potential mechanism of resistance to EGFR inhibition in RAS-wild type (WT) CRC. Little is known about molecular characteristics of NF1-mutated (MT) CRC. Methods: Tumor profiles from 8150 CRC patients (pts) with available NF1 mutation status were retrospectively reviewed. NextGen sequencing by a customized 592-gene panel was performed. Microsatellite instability (MSI) / mismatch repair (MMR) status, tumor mutational burden (TMB) and PD-L1 expression were tested. Molecular profiles between NF1-MT and NF1-WT pts were compared. Results: Out of 8150 pts, 176 (2.2%) had somatic NF1 mutations with pathogenic or presumed pathogenic function. A higher NF1-MT frequency was observed in MSI-H/dMMR vs MSS/pMMR (13.5% vs 1.4%, p < 0.0001), in right-sided vs left sided (2.9% vs 1.8%, p < 0.01), and in RAS-WT vs RAS-MT (3.0% vs 1.4%, p < 0.0001). In MSS/pMMR tumors, no association with sidedness was observed (right: 1.3% vs left: 1.2%, NS). The most prevalent co-mutations with NF-1 were APC (63.2%), ARID1A (57.5%), TP53 (51.5%), KMT2D (32.9%) and KRAS (32.4%) in all cases, and APC (76.2%), TP53 (69.5%), KRAS (38.8%), ARID1A (34.4%) and FBXW7 (21.5%) in MSS/pMMR cases. POLE mutation was observed in 18.4% of NF1-MT/MSS/pMMR pts. Compared to NF1-WT pts, NF1-MT pts had more frequent mutations in ARID1A (All: 57.5% vs 23.3%, p < 0.0001; MSS/pMMR: 34.4% vs 15.2%, p < 0.05), and less frequent mutations in KRAS (All: 32.4% vs 49.0%, p < 0.0001; MSS/pMMR: 38.8% vs 50.3%, p < 0.05). Also, NF1-MT pts had more frequent alterations in homologous recombination pathway compared to NF1-WT pts (All: 39.8% vs 7.5%, p < 0.0001; MSS/pMMR: 17.5% vs 4.4%, p < 0.0001). Mean TMB was significantly greater in NF1-MT than NF1-WT (All: 48.9/Mb vs 10.0/Mb, p < 0.0001; MSS/pMMR: 48.3/Mb vs 8.2/Mb, p < 0.0001). Also, PD-L1 positivity was higher in NF1-MT compared to NF1-WT (All: 12.9% vs 3.6%, p < 0.0001; MSS/pMMR: 7.1% vs 2.6%, p < 0.05). Conclusions: While more frequent than in RAS-MT pts, NF1-MT CRC was a small subset in RAS-WT pts. NF1-MT was associated with alterations in chromatin remodeling and DNA damage response pathways, as well as elevated TMB and PD-L1 expression, which may provide alternative therapeutic strategies beyond EGFR inhibition.

Comprehensive molecular analysis of microsatellite-stable (MSS) tumors with high mutational burden in gastrointestinal (GI) cancers

3631

Background: Mutational signatures contributing to high tumor mutation burden (TMB-H) independent from microsatellite instability-high (MSI-H) status are not well-studied. We aimed to characterize specific molecular features of a large cohort of GI tumors with TMB-H & MSS. Methods: We sequenced23392 GI tumors, including 2707 gastroesophageal (GE), 11616 colorectal (CRC), and 9069 others. Samples were analyzed using Next-generation sequencing (NGS) and immunohistochemistry (IHC) (Caris Life Sciences, Phoenix, AZ). MMR/MSI status was evaluated by a combination of IHC, Fragment Analysis and NGS. Tumors with TMB ≥ 17 mutations/Mb were defined as TMB-H. PD-L1 was tested by IHC [22C3 (CPS score, positivity: CPS ≥ 1%) in GE tumors and SP142 (Positivity: TPS ≥ 5%) in other cancers]. Findings were compared in four groups (TMB-H/L & MSI-H/MSS) using Fisher-Exact or Chi-square and adjusted for multiple comparison by Benjamini-Hochberg. Significance was determined by adjusted (adj) p < .05. Results: Overall, TMB-H & MSS was observed in 1% of patients (pts) (n = 237, including 45 GE, 124 CRC, 68 others), while TMB-H & MSI-H, TMB-L & MSS, TMB-L & MSI-H were observed in 4% (n = 936), 94.4% (n = 22089) and 0.6% (n = 130) respectively. Compared to other groups, TMB-H & MSS showed the most prevalent amplifications (AMPs), including CCND1 (5.6%), FGF3/4/19 (4.9%, 4.3%, 4.4%) , MYC (4.3%) (Top 5, adj p < .05), and the highest mutation rates in POLE (21.6%), RB1 (13.1%), CDC73 (10.3%), RUNX1 (6.5%), and genes involved in PI3K & MAPK ( PIK3R1 17%, mTOR 3.4%, MAP2K1 3.8% , MAP2K4 5.6%) and Wnt ( APC 48.5% , SMAD2 6.5% , TCF7L2 3.8%) pathways (adj p < .05). The rates of HER2 AMP and high-expression (IHC) were the highest in TMB-H & MSS, followed by TMB-L & MSS, TMB-H & MSI-H, TMB-L & MSI-H (adj p< .0001); PD-L1 positive rate was similar between TMB-H & MSS and TMB-L & MSI-H, while the highest and lowest in TMB-H & MSI-H and TMB-L & MSS respectively in GE and other GI cancers (adj p < .01) (Table). Conclusions: This is the largest study to investigate the special molecular landscape of pts with TMB-H & MSS in GI cancers. Our data may provide novel insights for pt selection and more effective targeted combination immunotherapies (e.g. HER2, PI3K inhibitors) in GI cancers. [Table: see text]

The landscape of DNA damage response (DDR) pathway in colorectal cancer (CRC)

4064

Background: Abnormal DDR is a hallmark of cancer, relating to genome instability, anti-tumor immunity, and sensitivity to chemotherapeutic agents and radiation. We conducted a large-scale investigation to clarify the alteration of DDR pathway in CRC. Methods: Tumor samples from 9321 CRC patients were retrospectively reviewed. Next-Generation Sequencing (NGS) on a custom-designed panel enriching 592 gene targets was performed. Samples with mutations detected in any of 29 DDR-related genes were deemed DDR-mutant (DDR-MT); the rest DDR-wild type (DDR-WT). Microsatellite instability (MSI) status was tested with a combination of immunohistochemistry (IHC), fragment analysis and NGS. Tumor mutational burden (TMB) was calculated based on somatic nonsynonymous missense mutations. PD-L1 was tested by IHC (SP142). Consensus molecular subtype (CMS) was developed using RNA sequencing data. Results: Of 9321 cases, 1290 (13.8%) were DDR-MT. DDR-MT frequency was higher in right vs. left sided (20.9% vs 10.8%, p < 0.001) and MSI-H vs. MSS (76.4% vs 9.5%, p < 0.001) cases. In the MSS cases, right-sided had marginally higher frequency of DDR-MT than left-sided (10.6% vs 9.1%, p = 0.055), with much higher frequency of Fanconi anemia pathway alteration in right-sided (1.5% vs 0.7%, p < 0.01). CMS1 subtype had the highest frequency of DDR-MT (34.8%); CMS2 had the lowest (7.1%). DDR-MT cases (vs. DDR-WT) had higher mutation rate of ARID1A (55.0% vs 19.1%, p < 0.0001), PIK3CA (22.6% vs 15.8%, p < 0.0001) and BRAF (20.4% vs 7.3%, p < 0.0001), and lower mutation rate of TP53 (48.2% vs 76.1%, p < 0.0001), APC (60.5% vs 74.5%, p < 0.0001) and KRAS (44.0% vs 49.8%, p < 0.001). Mean TMB was much greater in DDR-MT than DDR-WT (All: 20.9/Mb vs 7.7/Mb, p < 0.0001; MSS: 13.7/Mb vs 7.6/Mb, p < 0.05). PD-L1 positivity was also higher in DDR-MT compared to DDR-WT (All: 10.1% vs 2.7%, p < 0.0001; MSS: 4.8% vs 2.4%, p < 0.0001). Conclusions: Alteration of the DDR pathway was strongly associated with MSI status in CRC. The primary tumor sidedness might also be related, as DDR-MT was more prevalent in right-sided tumors. Elevated TMB and PD-L1 expression in DDR-MT CRC indicate more activated anti-tumor immune profiles compared to DDR-WT, regardless of MSI status, suggesting possible therapeutic benefit from immune checkpoint inhibitors in DDR-MT CRC.

SWOG S1505: Results of perioperative chemotherapy (peri-op CTx) with mfolfirinox versus gemcitabine/nab-paclitaxel (Gem/nabP) for resectable pancreatic ductal adenocarcinoma (PDA)

4504

Background: Clinical outcomes after curative treatment of resectable PDA remain suboptimal. To assess the potential of early control of systemic disease with multiagent peri-op CTx, we conducted a prospective trial in the National Clinical Trials Network. Methods: S1505 was a randomized phase II trial of peri-op CTx (12 weeks pre-, 12 weeks post-op) with either mFOLFIRINOX (Arm 1) or Gem/nabP (Arm 2). Eligibility required confirmed tissue diagnosis of PDA, ECOG PS 0 or 1, and resectable disease per Intergroup criteria. Primary outcome was 2-year overall survival (OS), using a “pick the winner” design; for 100 eligible patients (pts), accrual up to 150 pts was planned to account for cases deemed ineligible at central radiology review. We previously presented data on eligibility (ASCO 2019 abstr 4137). Here we present the final efficacy and toxicity results for the eligible pts. Results: From 2015 to 2018, 147 pts were enrolled; there were 102 eligible pts; 55 in Arm 1; 47 in Arm 2. For Arms 1 and 2 respectively: median age, 66 (44-76) and 64 (46-76) years; males, 36 (65%) and 24 (51%); and ECOG PS 0, 34 (62%) and 31 (66%) pts. Treatment details are shown in Table. For Arm 1 and Arm 2, respectively: Two-year OS was 41.6% and 48.8%; median OS was 22.4 months and 23.6 months. Neither arm’s 2-year OS estimate was statistically significantly higher than the a priori threshold of 40% (p=0.42 in Arm 1 and p=0.12 in Arm 2). Median disease-free survival (DFS) after resection was 10.9 months in Arm 1 and 14.2 months in Arm 2 (p=0.87). Conclusions: We have demonstrated: 1) two-year OS of 41.6% (median 22.4 months) with mFOLFIRINOX and 48.8% (median 23.6 months) with Gem/nabP for all eligible pts starting treatment for resectable PDA; 2) post-resection DFS of 10.9 months and 14.2 months, respectively; 3) adequate safety and high resectability rates with peri-op CTx; 4) little evidence that either regimen improves OS compared with the historical standard. Clinical trial information: NCT02562716 . [Table: see text]

Profiling and prognostic implications of variants of unknown significance (VUS) in solid tumors

e19012

Background: While precision oncology is becoming increasingly integrated into standard of care for most incurable solid tumor malignancies, there is paucity of data regarding the clinical significance of VUS. In this study, we aim to evaluate whether the number of VUS is associated with overall survival (OS). We also analyze racial disparities pertaining to the number of VUS present and identify which, if any, genes possess prognostic implications. Methods: This is a retrospective review of 389 consecutive patients seen at Cleveland Clinic from 2014 to 2016 with incurable solid tumor malignancies, for whom next-generation sequencing (NGS) was ordered using Foundation One™ (Cambridge, MA). Demographics, number of VUS, genes involved, and race were summarized. OS was estimated by Kaplan-Meier and compared by log rank test. Results: Median age was 60 years, 202 (52%) patients were female, 338 (86.7%) were Caucasian, and 31 (8.0%) were African American. On NGS, 376 (97%) patients had VUS reported. The median number of VUS was 9 (range 1-116). When dichotomized at the median, the number of VUS did not affect OS. Genes most commonly implicated in reported VUS were LRP1B (88, 22.6%), MLL3 (83, 21.3%), MLL2 (73, 18.8%), ARID1B (70, 18.0%), PRKDC (60, 15.4%), PREX2 (58, 18.7%), and SPTA1 (56, 14.4%). Patients found to have a variant of unknown significance in MLL2 had worse median OS as compared to those who did not (2.61 vs 3.76 years respectively; p = 0.033). When profiled by race, Caucasians had lower numbers of VUS (p = 0.002; Table). Conclusions: We did not find a clear association between the number of VUS and OS. MLL2, a gene known to predict poor prognosis as a pathogenic variant, was seen in our study to have similarly poor prognostic implications as a variant of unknown significance. Racial disparities in genomics exist as African Americans are under-represented and have greater numbers of VUS as compared to Caucasians. Further research is warranted to elucidate these disparities. [Table: see text]

Real-world outcomes in pancreatic adenocarcinoma (PDAC) and persona types with implications for standard of care (SOC) therapy (Tx)

735

Background: Molecular profiling (MP) for PDAC has gained increased acceptance and we previously demonstrated that targeting actionable mutations can improve patient (pt) outcomes. However, the correlations of diverse patterns of molecular alterations with outcomes following SOC Tx are largely unknown. Methods: We analyzed longitudinal outcomes of 1355 PDAC pts who underwent MP and received SOC Tx. “Persona” types were established based on the molecular characteristics of each pt using unsupervised clustering, as well as a supervised review defined by our molecular tumor board, following classifications reported in previous studies. Progression-free survival (PFS) for each type was assessed based on the choice of first-line Tx (i.e. FOLFIRINOX [FFX] vs. gemcitabine + nab-paclitaxel [GA]). Statistical comparisons were made against all other types within a specific Tx group. Results: The prognostic/predictive value of the persona types for 1st-line Tx revealed distinct differences in outcomes (Table). As expected, the DDR deficiency type was associated with a significantly improved PFS for pts treated with FFX but not for GA. In addition, pts in the cell cycle type had a worse PFS compared to other persona types for both FFX and GA. Using this platform, we will further subdivide the persona types into molecular subtypes and associate these with pt outcomes. Conclusions: Our analyses demonstrate that specific molecular persona types exist in PDAC pts and can be linked to Tx outcomes. Ultimately, knowing the persona type/subtype early in a pt’s Tx course may help personalize Tx to improve outcomes. [Table: see text]

A pilot clinical trial of p53/p16-independent epigenetic therapy for pancreatic ductal adenocarcinoma (PDA)

699

Background: PDA treatment is limited to cytotoxic drugs. A key factor limiting their efficacy is TP53 mutations, omnipresent in PDA, which counter apoptosis-mediated cell kill. We evaluated a novel epigenetic approach using decitabine (Dec) to inhibit DNA methyltransferase 1 (DNMT1) and effect cancer cell cycle exits by epithelial-differentiation, combined with tetrahydrouridine (THU) to inhibit cytidine deaminase (CDA) and thereby permit oral bioavailability and solid-tissue distribution of Dec. Methods: Open-label single-arm, IRB-approved clinical trial at Cleveland Clinic and University Hospitals for patients with metastatic PDA that had progressed on prior chemotherapy, ECOG PS of 0-2. Treatment was oral, weight-based, with Dec 10-20 mg, and THU 500-1000 mg daily, 5 days/week. Primary endpoint was DNMT1 protein levels at 16-week vs baseline biopsies. Results: From Apr to Aug 2017, we enrolled 13 patients. Median age was 65 (range 44-74) years; 7 (54%) males; 11 (85%) Caucasians. Median time from diagnosis was 13 (3.9-53.5) months, with a median of 2 (1-3) prior lines of therapy. Baseline ECOG PS was 0/1 in 12 (92%) patients. All patients started study drugs; median time on treatment was 35 (4-63) days, and on study 72 (25-105) days. The most frequent adverse events attributable to the study drugs were anemia (n=5), and anorexia, dehydration, nausea, fatigue, febrile neutropenia and decreased lymphocyte count, in 3 patients each; no deaths. Eight (62%) patients underwent evaluation scans at 8 weeks, showing stable disease in 1 patient and progression in 7. Common reasons for coming off of study drugs were progression (n=6), physician discretion (n=3), and adverse events (n=2). Overall, 6 patients died; median survival was 3.1 months, and patients did not reach the 16-week biopsy. Shifts in blood counts, a sensitive indicator of Dec systemic activity, were unexpectedly mild, and plasma CDA enzyme activity was increased versus other cancer and normal controls. Conclusions: This first-of-its-kind study demonstrated feasibility and safety of the novel oral epigenetic therapy. Systemically elevated CDA in these patients requires higher doses of THU; a trial accordingly refined is planned. Clinical trial information: NCT02847000.

Abstract C58: p53/p16-independent cytoreduction of chemoresistant pancreatic adenocarcinoma by metabolically optimized epigenetic therapy

Although several reasons for chemorefractoriness of pancreatic ductal adenocarcinoma (PDAC) response have been proposed, a well-established basis is genetic inactivation of the p53/p16 apoptosis axis. Therefore, there is a need for p53/p16-independent PDAC treatments. The key epigenetic regulator DNA methyltransferase 1 (DNMT1) has been scientifically validated as a molecular target for p53-independent cytoreduction of pancreatic cancer, e.g., by activating terminal epithelial programs. Moreover, the pyrimidine nucleoside analog prodrugs decitabine (Dec) or 5-azacytidine (5Aza) are metabolically processed into a nucleotide analog that depletes DNMT1. Motivated by the scientific data, several clinical trials evaluating Dec, 5Aza, or analogs to treat PDAC have been completed. However, results from these trials have been discouraging and contradict encouraging preclinical results. To model the clinical situation, we evaluated Dec and 5Aza efficacy in gemcitabine-resistant PDAC in vitro and in mice. Gemcitabine-resistant PDAC cells were cross-resistant to Dec but not 5Aza. Measurement of the pyrimidine metabolism enzymes needed for prodrug conversion demonstrated the reason: DCK that activates gemcitabine and Dec was suppressed, but UCK2 that activates 5Aza was preserved and upregulated. Also, the catabolic enzyme cytidine deaminase (CDA), which rapidly catabolizes Dec and 5Aza into inactive uridine counterparts, was upregulated. Importantly, these shifts in pyrimidine metabolism that were observed in established gemcitabine-resistant cells were rapidly recapitulated (within hours/days) upon exposure of naïve PDAC cells to Dec or 5Aza, although 5Aza exposure immediately upregulated DCK, whereas Dec exposure immediately upregulated UCK2. Measurements of nucleotide levels by LC-MS/MS indicated that these contrasting shifts were caused by opposite effects of Dec and 5Aza on dCTP levels. These observations suggested potential solutions to overcome resistance: We evaluated alternating Dec with 5Aza, to exploit their mutual priming effects on DCK and UCK2 expression, and the addition of tetrahydrouridine (THU) to inhibit the catabolism of Dec and 5Aza by CDA. This optimized regimen was very efficacious in a murine xenograft model of gemcitabine-resistant PDAC (median vehicle control tumor measurements 972 mm3(range 726-1267.5); median THU-Dec/THU-5Aza 16 mm3 (range 0-87.5); P&lt;0.00001). Terminal epithelial-differentiation of gemcitabine-resistant PDAC cells was confirmed by a significant increase in pancreatic epithelial markers PTF1A, MIST1, and N-cadherin, while apoptosis markers were not induced. In sum, the pyrimidine metabolism network, which senses and regulates nucleotide amounts, automatically dampens Dec (and gemcitabine) and 5Aza activity. Nevertheless, these reactions can be anticipated and exploited instead to extend and deepen DNMT1 depletion and noncytotoxic tumor cytoreduction.

Citation Format: Rita Tohme, Xiaorong Gu, Asmaa El Desoky, Caroline Schuerger, Daniel Lindner, Davendra Sohal, Yogen Saunthararajah. p53/p16-independent cytoreduction of chemoresistant pancreatic adenocarcinoma by metabolically optimized epigenetic therapy [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2019 Sept 6-9; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2019;79(24 Suppl):Abstract nr C58.

Abstract A080: Olaparib and the ATR inhibitor AZD6738 in relapsed, refractory cancer patients with homologous recombination (HR) repair mutations – OLAPCO

Introduction: Olaparib is a PARP inhibitor (PARPi) that provides significant clinical benefit in several BRCA-mutant cancers, including ovarian, breast, pancreas and prostate. The benefit is reduced considerably in patients with multiple prior lines of therapy and olaparib resistance has no specific treatment. PARP inhibition results in replication stress (RS) due to unrepaired single strand DNA breaks (SSB) and PARP trapping in BRCA- and other HR repair-deficient tumors. ATR has critical roles in the cellular response to SSB and RS. This makes ATR inhibitors an attractive partner with olaparib, since ATR inhibition has the potential to reverse the two major mechanisms of PARP inhibitor resistance, including restored HR or stabilization of stalled replication forks. The OLAPCO trial (NCT02576444 clinicaltrials.gov) investigated. the combination of olaparib and AZD6738, an inhibitor of ATR, in relapsed, refractory cancer patients with tumors harboring HR repair mutations and in patients with BRCA-mutated PARPi pre-treated/resistant high-grade serous ovarian cancer (HGSOC). Methods: Patients with treatment-refractory, relapsed cancer were enrolled at 4 participating centers. Germline and somatic mutations had to be deleterious by COSMIC or ClinVar for eligibility. Performance status and organ function requirements were standard for early phase trials. Olaparib was given at 300 mg bid daily and AZD6738 at 160 mg daily days 1-7 in a 28-day cycle. Patients were treated until progression. Objective response was assessed by RECIST1.1 and toxicity was assessed by CTCAE4.0. Endpoints were confirmed complete [CR] or partial [PR] response rate and clinical benefit (CB) rate (response [PR] and stable disease [SD] for &gt; 16 weeks). Results: We enrolled 24 patients; 17 (71%) females; median age 59 (36-78) years. Patients were heavily pretreated, with a median of 4 (0-10) prior regimens. Myelosuppression, especially anemia and thrombocytopenia, was the most frequent toxicity but no patient required discontinuation. Two patients required olaparib reductions for anemia. At the time of data cut-off, 20 patients are evaluable for response assessment. One of 5 patients with ATM mutations had a CR, 2 patients have CB ongoing at 12+ months. Of 7 patients with HGSOC resistant to platinum and PARP inhibitors (1-3 prior agents), 1 achieved a PR ( -90%), 3 had a best response of SD with regression &lt; 30% (1 ongoing at 1 year) and 3 patients had progression (PD) as best response (Table 1). No other mutation or cancer type demonstrated objective response. Conclusions: The combination of olaparib and the ATR inhibitor AZD6738 demonstrated preliminary activity in patients with tumors harboring ATM mutations and in PARPi-resistant BRCA1/2-mutated HGSOC. Activity in ATM loss with an ATR inhibitor is consistent with the expected synthetic lethality of these interwoven DNA repair pathways. The encouraging data in HGSOC patients who have already progressed on a PARP inhibitor warrants additional study to further define the potential of this regimen to reverse PARPi resistance in BRCA-mutated HGSOC and other relevant solid tumors. The durability of responses in both groups (4 &gt;1 year) is promising. Table 1MutationATMBRCA prostate pancreasBRCA Prior PARPi HGSOCCHEK2MUS81PALB2IDH and SDHD #5471142 CR1 PR0 1 SD &gt; 4 mos213 11 PD 33112 Inevaluable110 11

Citation Format: Joseph Paul Eder, Davendra Sohal, Haider Mahdi, Khanh Do, Vicki Keedy, Navid Hafez, Deborah Doroshow, Manuel Avedissian, Peter Mortimer, Colin Glover, Patricia LoRusso, Juliane M Juergensmeier, Geoffrey I Shapiro. Olaparib and the ATR inhibitor AZD6738 in relapsed, refractory cancer patients with homologous recombination (HR) repair mutations – OLAPCO [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 A080. doi:10.1158/1535-7163.TARG-19-A080

Lactate-mediated epigenetic reprogramming regulates formation of human pancreatic cancer-associated fibroblasts

Even though pancreatic ductal adenocarcinoma (PDAC) is associated with fibrotic stroma, the molecular pathways regulating the formation of cancer associated fibroblasts (CAFs) are not well elucidated. An epigenomic analysis of patient-derived and de-novo generated CAFs demonstrated widespread loss of cytosine methylation that was associated with overexpression of various inflammatory transcripts including CXCR4. Co-culture of neoplastic cells with CAFs led to increased invasiveness that was abrogated by inhibition of CXCR4. Metabolite tracing revealed that lactate produced by neoplastic cells leads to increased production of alpha-ketoglutarate (aKG) within mesenchymal stem cells (MSCs). In turn, aKG mediated activation of the demethylase TET enzyme led to decreased cytosine methylation and increased hydroxymethylation during de novo differentiation of MSCs to CAF. Co-injection of neoplastic cells with TET-deficient MSCs inhibited tumor growth in vivo. Thus, in PDAC, a tumor-mediated lactate flux is associated with widespread epigenomic reprogramming that is seen during CAF formation.

SWOG S1505: Initial findings on eligibility and neoadjuvant chemotherapy experience with mFOLFIRINOX versus gemcitabine/nab-paclitaxel for resectable pancreatic adenocarcinoma

4137

Background: Clinical outcomes after curative therapy of resectable pancreatic ductal adenocarcinoma (PDA) remain suboptimal. For early control of systemic disease with aggressive perioperative chemotherapy (CTx), we conducted a prospective trial in the National Clinical Trials Network (NCTN) setting. Methods: S1505 was a randomized phase II trial of periop (12 weeks pre-, 12 weeks post-op) CTx with either mFOLFIRINOX (5-fluorouracil, irinotecan, oxaliplatin – without bolus 5-FU and leucovorin; Arm 1), or gemcitabine/nab-paclitaxel (Arm 2). Eligibility required adult patients with ECOG PS 0 or 1, confirmed tissue diagnosis of PDA, and resectable disease: no involvement of the celiac, common hepatic, or superior mesenteric arteries (and, if present, variants); < 180° interface between tumor and vessel wall, of the portal or superior mesenteric veins; patent portal vein/splenic vein confluence; no metastases. Primary outcome is 2-year overall survival (OS), using a “pick the winner” design; for 100 eligible patients, accrual up to 150 patients was planned, to account for cases deemed ineligible at central radiology review. Results: From 2015 to 2018, 147 patients were enrolled; 74 to Arm 1; 73 to Arm 2. At central radiology review, 42/147 (29%) were ineligible; of these, 15 (36%) had venous involvement ≥180°, 22 (52%) had arterial involvement, 28 (67%) had distant disease. One patient had distal cholangiocarcinoma (ineligible); one withdrew consent after randomization. Eligible patients (n = 103) had median age 64 years; males 58%; whites 89%; PS 0 64%. Of 103, 99 (96%) started and 86 (83%) completed preop CTx. There was one death due to sepsis and 61 additional patients experienced grade 3/4 toxicities. To date, 76 of 99 (77%) patients went to surgery and 72 (73%) underwent resection. Conclusions: This is the first-ever NCTN study of periop CTx for resectable PDA. Accrual was brisk, establishing feasibility. Ineligible cases after central radiology review highlight quality control and physician education imperatives for neoadjuvant PDA trials. Preop CTx safety and resection rates are encouraging. Follow up for OS is ongoing. Clinical trial information: NCT02562716.

Accelerating advanced precision medicine through a harmonized data exchange platform and research consortium (PMEC)

6557

Background: Clinico-genomic data sharing is consistently identified by the global oncology community as a critical requirement to accelerate the discovery and development of new targeted therapies. However, lack of effective collaborative models, fragmented and lengthy legal contracting processes, paucity of funding, and inadequate technological platforms have historically been obstacles for effective data sharing. Methods: In 2015, 10 US academic medical centers (AMC) and Foundation Medicine Inc. (FMI) formed PMEC. Feasibility assessments included creation of a master agreement across sites and willingness to use a central IRB. Oversight and research steering committees were created within the consortium. Through a centralized, secure web-based platform, FoundationInsight, we combined and shared de-identified, harmonized comprehensive FoundationOne genomic profiling data. Research proposals mining this data warehouse are invited quarterly from participant AMCs and peer-reviewed; approved studies are executed at all sites. Results: All 10 AMCs collaborated to execute a master registry participation agreement, followed by a master IRB protocol (New England IRB # 120180008), subsequently approved by individual site IRBs. Since its launch, the PMEC database has grown, on average, 60% per year, to now house over 14,000 cases. The shared dataset covers all tumor types (most commonly lung [17.2%], gastrointestinal [13.8%] and breast [9.2%]), encompasses genomic alterations in >300 genes, and reports relevant supplementary data such as tumor mutation burden and microsatellite instability status. To date, 15 studies have been proposed and evaluated using this platform, with 2 projects currently approved and in progress. Conclusions: We demonstrated the feasibility of creating a collaborative academic consortium that facilitates data sharing and potential discovery efforts in oncology. Technology solutions can accelerate the ability of AMCs, in partnership with central labs, to share and harmonize data to advance precision medicine. This approach lays the groundwork for conducting prospective, biomarker-enriched clinical trials among participating AMCs.

Improved overall survival (OS) for advanced pancreatic cancer (PDAC) patients (pts) enrolled in the Know Your Tumor (KYT) program whose tumors harbored highly actionable molecular alterations and who received molecularly-matched therapies (tx)

4138

Background: Initial results from the KYT program demonstrated that 27% of PDACs harbor highly actionable molecular alterations (herein labelled “actionable biomarkers”), defined as biomarkers that predict for a high response rate to appropriately targeted tx, in any cancer type. Within this cohort, the median progression-free survival on molecularly-matched tx was 2 months longer than unmatched tx. Here, we present OS data emphasizing the 125 pts with “actionable biomarkers” who did or did not receive molecularly-matched tx. Methods: PanCAN and Perthera have coordinated tumor molecular profiling through commercial labs (NGS/IHC panels) for PDAC pts since 2014. Results are reviewed by a molecular tumor board, and tx options are prioritized based on the actionable biomarkers, in the context of the pt’s tx history. Pts are followed longitudinally to track physician tx choices and survival outcomes. Cox regression was used to assess differences in OS (measured from date of diagnosis until death). Results: Of 1053 pts who received a Perthera Report, 25% had “actionable biomarkers”. OS analyses across 454 pts with adequate tx history are shown in the Table below. Notably, pts with “actionable biomarkers” who received a molecularly-matched tx had a significantly increased OS compared to those with “actionable biomarkers” but who did not receive molecularly-matched tx. Subgroup analyses related to tx history and specific molecular pathways that warrant further investigation will be discussed. Conclusions: When the ~25% of PDAC pts whose tumors harbored “actionable biomarkers” received molecularly-matched tx, they had a better OS. These findings support the need to test all pts with PDAC, and just as importantly, to maximize access to molecularly-matched tx for appropriate pts, to achieve the best pt outcomes. [Table: see text]

Activity of larotrectinib in TRK fusion cancer patients with brain metastases or primary central nervous system tumors

2006

Background: TRK fusions are oncogenic drivers of a variety of cancers, many of which can involve the central nervous system (CNS). Larotrectinib is an FDA-approved selective TRK inhibitor for the treatment of TRK fusion cancer (Drilon et al, NEJM 2018). While larotrectinib has been shown to cross the blood–brain barrier (Ziegler et al, Br J Cancer 2018), its clinical activity in a series of TRK fusion cancers with primary or metastatic intracranial disease has not been described. Methods: Patients (pts) with non-primary CNS solid tumors with brain metastases, or primary CNS tumors harboring a TRK fusion treated with larotrectinib in 2 clinical trials (NCT02637687, NCT02576431) were identified. Larotrectinib was administered until disease progression (PD), withdrawal, or unacceptable toxicity. Disease status was investigator-assessed (RANO and RECIST). Data cutoff: July 30, 2018. Results: 14 pts were identified: 5 non-primary CNS solid tumors (3 lung cancer, 2 thyroid cancer; fusion type: 2 ETV6-NTRK3, 2 SQSTM1-NTRK3, 1 EPS15-NTRK1; age range 25–79 y) and 9 primary CNS tumors (3 glioma, 2 glioblastoma, 1 astrocytoma, 3 NOS; fusion type: 3 BCR-NTRK2, 2 KANK-NTRK2, 1 each of AFAP1-NTRK1, AGTPBP1-NTRK2, ETV6-NTRK3, SPECC1L-NTRK2; age range 2–79 y). In the 5 pts with non-primary CNS tumors, the best objective response to therapy was PR in 3 (60%, 1 pending confirmation), SD in 1 (20%), and not evaluable (NE) in 1 (20%). Duration of response ranged from 9+ to 13 mo. In the 9 pts with primary CNS tumors, disease control was achieved in all evaluable pts (primary PD not observed; 1 pt required dose increase). The best objective response to therapy was PR in 1 (11%; pending confirmation, −55% tumor shrinkage, ongoing at 3.7 mo), SD in 7 (78%; tumor shrinkage range −1% to −24% for pts with measurable disease, 5 had SD > 4 mo), and NE in 1 (11%). Duration of treatment ranged from 2.8–9.2+ mo. Conclusions: Larotrectinib is active in pts with TRK fusion cancers with intracranial disease. Confirmed responses and durable disease control were seen in metastatic disease and primary CNS tumors of various histologies. These results further support expanded testing for TRK fusions across all cancers, including primary CNS tumors. Clinical trial information: NCT02637687 and NCT02576431.

Microbiome signature of bile from pancreatic and biliary tract cancer patients: A pilot study

e15744

Background: Recent studies especially in murine models have linked meta-organismal pathways in the gut microbial community to cancers of the pancreas and the biliary tract. However, data on microbiome in the biliary pool in patients to establish models for oncogenesis in the pancreatobiliary (PB) habitat have been limited. We analyzed bile collected from a pilot series of patients to identify microbiome signatures associated with malignancy. Methods: We collected bile samples from patients during routine endoscopic retrograde cholangiopancreatography in this study approved by the Cleveland Clinic Institutional Review Board for Human Subjects’ Protection. Of 10 patients, there were 5 with pancreatic ductal adenocarcinoma (PDA), 3 with cholangiocarcinoma (CC), 1 with ampullary adenocarcinoma, and 1 with gallstone pancreatitis. DNA was extracted from bile specimens using PowerViral RNA/DNA Isolation kit. Bacterial 16S rRNA gene amplification and library construction were performed according to the 16S Metagenomic Sequencing Library Preparation guide from Illumina. Post-sequencing analysis was done using QIIME (Quantitative Insights Into Microbial Ecology) and MICCA (MICrobial Community Analysis). Results: Most reads were from phyla Firmicutes (57.9%) and Proteobacteria (14.9%). One benign specimen (pancreatitis) separated clearly from the rest showing 98.9% of reads from Clostridium sensu stricto (phylum Firmicutes). Analysis of beta diversity showed six samples clustering tightly. Of these, 5 were PDA and 1 was CC. A second cluster included remaining 3 samples, 2 with CC and 1 with PDA; this latter had higher abundance of phyla Fusobacteria (90.6%) and Verrucomicrobia (92.9%). Conclusions: Select bacteria are differentially increased in malignant and benign PB diseases. Furthermore, distinct microbiome signatures may be associated with cancer in the pancreatic and biliary habitats. We intend to evaluate these findings in larger sample sizes and determine associations with clinical outcomes and response to treatment.

SWOG S1505: Initial findings on eligibility and neoadjuvant chemotherapy experience with mfolfirinox versus gemcitabine/nab-paclitaxel for resectable pancreatic adenocarcinoma

414

Background: Clinical outcomes after curative therapy of resectable pancreatic ductal adenocarcinoma (PDA) remain suboptimal. For early control of systemic disease with aggressive perioperative chemotherapy (CTx), we conducted a prospective trial in the National Clinical Trials Network (NCTN) setting. Methods: S1505 was a randomized phase II trial of periop (12 weeks pre-, 12 weeks post-op) CTx with either mFOLFIRINOX (5-fluorouracil, irinotecan, oxaliplatin – without bolus 5-FU and leucovorin; Arm 1), or gemcitabine/nab-paclitaxel (Arm 2). Eligibility required adult patients with ECOG PS 0 or 1, confirmed tissue diagnosis of PDA, and resectable disease: no involvement of the celiac, common hepatic, or superior mesenteric arteries (and, if present, variants); < 180° interface between tumor and vessel wall, of the portal or superior mesenteric veins; patent portal vein/splenic vein confluence; no metastases. Primary outcome is 2-year overall survival (OS), using a “pick the winner” design; for 100 eligible patients, accrual up to 150 patients was planned, to account for cases deemed ineligible at central radiology review. Results: From 2015 to 2018, 147 patients were enrolled; 74 to Arm 1; 73 to Arm 2. At central radiology review, 42/147 (29%) were ineligible; of these, 15 (36%) had venous involvement ≥ 180°, 22 (52%) had arterial involvement, 28 (67%) had distant disease. One patient had distal cholangiocarcinoma (ineligible); one withdrew consent after randomization. Eligible patients (n = 103) had median age 64 years; males 58%; whites 89%; PS 0 64%. Of 103, 99 (96%) started and 86 (83%) completed preop CTx. There was one death due to sepsis and 61 additional patients experienced grade 3/4 toxicities. To date, 76 of 99 (77%) patients went to surgery and 72 (73%) underwent resection. Conclusions: This is the first-ever NCTN study of periop CTx for resectable PDA. Accrual was brisk, establishing feasibility. Ineligible cases after central radiology review highlight quality control and physician education imperatives for neoadjuvant PDA trials. Preop CTx safety and resection rates are encouraging. Follow up for OS is ongoing. Clinical trial information: NCT02562716.