ALBERT

Description: AML patients (pts) with a normal chromosome pattern and the “Internal Tandem Duplication” (ITD) of the FLT3 gene have an overall and event-free survival (OS, EFS) inferior than those of pts with a FLT3 “wild-type” gene because of a higher relapse risk, are placed in the intermediate-1 prognostic category, and when in complete remission (CR) are candidates to allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, the role of allo-HSCT in ITD+ pts is still debated and a recent retrospective analysis has stated that FLT3/ITD adversely affected allo-HSCT outcome in the same direction as it does after chemotherapy. Thus, the present study was aimed to compare the OS, EFS and relapse risk of pts submitted to allo-HSCT in first CR or in initial relapse, defined by a 5-10% bone marrow blast cell percentage, with those of pts submitted to chemotherapy alone. The study cohort consisted of 54 chromosomally normal, ITD+ consecutive non-M3 AML pts who were included in 168 AML pts aged 18-66 years who came to our observation in the period January 2007-December 2013. At diagnosis median age was 48.6 years (range 18-66), 25 pts were males and 29 females. Median follow-up was 16.2 months (0.4-68.8): median follow-up for responsive pts was 7.15 months (range 0.26-27.6), for relapsed pts was 18.1 (range 1.9-68.8) and for transplanted pts was 9.1 months (range 2.9-26.8). All pts received the same induction treatment that consisted of standard Idarubicine+Ara-c “3+7” followed by two consolidation courses with high-dose Ara-c. Those who failed induction received other treatment schedules among which Fludarabine+Ara-c+Idarubicine was the most common. At the end of consolidation 38 pts were in first CR, achieved after 3+7 in 27 pts and after a second different course in 11. Sixteen pts had a resistant disease. After a median time of five months (range 1-21) 21/38 pts (55.2%) relapsed. All received a re-induction and 8/21 (38.1%) attained a second CR. Allo-HSCT was performed in a total of 23 pts: 12 first CRs, 6 second CRs and 5 initial relapses. The donor was a sibling in 7 pts, an unrelated donor in 13 and an haplo-identical donor in 3; the HSC source was the marrow in 6 pts, the peripheral blood in 16 and the cord blood (CB) in one. The Conditioning regimen was myeloblative (mainly Busulfan+Fludarabine) in 21 and non-myeloblative in 2; GvHD prophylaxis consisted of Cyclosporine A, steroids and methotrexate “short course”. The median number of CD34+ cells infused was 5.14x106/kg (1.3-12.7). All pts except that who received CB engrafted after a median time of 15 days (12-28); 21 were complete chimeras, two partial chimeras. Thirteen pts developed acute GvHD (grade I in 3 pts, grade II in 5, grade III in 2 and grade IV in 3) which totally/partially recovered after high dose steroids along with different immunosuppressive drugs in 4 and 9 pts respectively. Eleven pts developed a chronic GvHD which was the evolution of an aGvHD in 9, targeted different organs, was stable in 3 pts and completely/partially recovered in 4 and 2 pts respectively. Post-transplant relapse occurred in 3/12 first CRs, in none second CRs and in 3/5 initial relapse. The estimated response rate for CR pts submitted to allo-HSCT was 422.1 (95% CI: 262.4-679.1) versus 64.6 (95% CI: 40.7-102.6) for those submitted to chemotherapy alone with a median survival time of 7.2 months (range 5.3-8.8) versus not reached (1.9-not available); on univariable Cox model the HR of allo-HSCT pts was 37.9 (95% CI: 9.4-152.0) with p=0.0000. The estimated relapse rate for CR pts submitted to allo-HSCT was 8.6 (95% CI: 2.1-34.4) versus 44.3 (95% CI: 25.7-76.3) for those submitted to chemotherapy alone with a median survival time not reached (range not available) versus 13.2 (7.2-not available); on univariable Cox model the HR of allo-HSCT pts was 0.5 (95% CI: 0.2-1.1) with p=0.06. The estimated death rate for CR pts submitted to allo-HSCT was 28.7 (95% CI: 13.7-60.4) versus 49.7 (95% CI: 32.1-77.1) for those submitted to chemotherapy alone with a median survival time of 18.3 months (range 14.2-not available) versus 12.2 (8.8-18.9); on univariable Cox model the HR for allo-HSCT pts was 0.48 (95% CI: 0.2-1.1) with p=0.09. In conclusion, our series suggests that in ITD+ pts allo-HSCT significantly strengthens CR in pts who had already responded to conventional chemotherapy, but it presents only a trend towards significance when its superiority to prevent relapse was considered. Disclosures Castagnola: Gilead Sciences: Research Funding.

Description: Introduction: Postremission treatment in adult acute myeloid leukemia (AML) is mandatory to improve long-term remission duration. Allogeneic stem-cell transplantation (allo-SCT) is considered gold standard in younger patients (pts) with higher risk of relapse, whereas in pts categorized as standard-risk (SR) AML the optimal treatment remains an open issue. High-dose cytarabine (HD-ARAC) is considered the backbone of postremission treatment in SR AML. The optimal dose and schedule, and the clinical benefit of additional chemotherapeutic agents remain controversial (Schaich, 2011- Lowenberg, 2011- Lowenberg, 2013) as well as the relative efficacy of HD-ARAC in the different cytogenetic and molecular subgroups of SR AML. We analyzed the very long term outcome of consecutive unselected AML pts treated according to AML 00 NILG protocol (Bassan, Blood, 2003: 102; 11) with three repetitive HD-ARAC-based cycles, associated with idarubicin, followed by limited peripheral blood stem cell (PBSC) support, in order to reduce toxicities. Methods: We evaluated 263 consecutive AML pts, diagnosed at our Institution, from January 2001 to March 2016 (median age 52 years - range 15-65). The treatment plan included induction with ICE (Idarubicin-Cytarabine-Etoposide) followed by IC consolidation or HD-ARAC/HD-idarubicin (SPLIT therapy) in refractory cases. ARAC 1g/mq bd for 4 days (A8) was given to CR pts, to collect 3-6x10^6 CD34/kg in 3 aliquots. Patients with ELN favorable or intermediate risk AML, considered SR AML, received 3 cycles of HD-ARAC (2g/mq bd for 5 days) plus idarubicin (8mg/mq for 2 days) with PBSC reinfusion (1-2x10^6 CD34/Kg) (A20). The pts with insufficient CD34 cells yield, received 2 courses HD-ARAC (1g/mq bd for 5 days) plus idarubicin (10mg/mq day 1) (A10). The cumulative dose of ARAC and idarubicin given according to A20 protocol were 69,4 g/mq and 162 mg/mq, respectively. Results: Among 263 pts, 142 (54%) pts were considered SR and 116 (44.1%) HR. Only 5 pts were lacking molecular or cytogenetic information (1.9%) to allow ELN risk classification. Accordingly, 65 pts (25.2%) were favorable [35.4% core binding factor (CBF) positive, 9.2% CEPBA mutated and 55.4% NPM mutated], 87 (33.7%) Intermediate-I, 45 (17.4%) Intermediate II and 61 (23.7%) adverse risk. CR rate after ICE treatment was 77.6% (204/263), 95.8% in SR and 56% in HR pts and after ICE+SPLIT it was 88.2% (232/263); early or aplastic death was 6%. A8 course was administered in 218 pts, of whom 179 were successfully mobilized (82.1%). Among 136 SR pts, 127 were received HD-ARAC (A20 or A10), 4 allo-SCT for early relapse, 3 pts did not proceed to A20 because of prolonged cytopenia and 2 are ongoing. Hematological toxicity was acceptable. ANC recovered at a median of 10 days after CD34+ cells reinfusion. Two death in CR occurred within 100 days after A20 therapy (1.6%). After a median follow-up of 53.3 months (range 1-172), the median survival of SR AML patients was 118.9 months, the DFS and OS at 5 y were 51+/-4.8% and 56.5+/-4.7%, and at 10 y 43.3+/-5.9% and 49.6+/-5.7%, respectively. According to ELN risk, the OS at 5 y and 10 y was 78.3+/-6% and 68+/-8.7% in favorable (median OS not reached), 52.3+/-8% and 46+/-9.4% in Intermediate-I (median OS 118.8 months), and 42+10% and 33.6+/-10.9% in Intermediate-II (median OS 33.1 months) (p0.0034) (Figure1). According to molecular characteristics, the OS at 5-y was 100% in CEBPA-mutated, 77+/-9% (+/-SE) in CBF-mutated and 71+/-11% in NPM-mutated subgroups, respectively (Figure 2). In NPM-mutated pts it was 56+/-14% if FLT3 was mutated and 75+/-10% if not. At 8-y OS was not evaluable in CEBPA-mutated, it remained 77+/-9% in CBF mutated but dropped to 47+/-15% in NPM-mutated pts due to late relapses (28%), which occurred in pts without FLT3. The median time to relapse in NPM pts was 17 m, the latest relapse occurred after 97 months. Conclusion: A postremission program with repeated courses of HD-ARAC associated to idarubicin and limited autologous CD34+ cells support was feasible, well tolerated and very effective, also long-term, in pts with ELN favourable or intermediate risk AML. The late relapse tendency of NPM-mutated AML is puzzling and should be considered by innovative programs aimed at improving these results. Figure 1 Figure 1. Figure 2 Figure 2. Disclosures No relevant conflicts of interest to declare.

Description: Introduction In AML and MDS cases, the genetic lesions inherited or acquired by the hematopoietic stem cells are considered as starting events. Familial AML and MDS, recently recognized in the revised WHO classification (2016) provide a useful model for investigation of predisposing genetic mutations. Genetic analysis of several pure familial leukemia pedigrees led to the discovery of well defined syndromes associated with inherited de novo mutations on germline DNA. Growing clinical awareness as well as a widespread use of NGS have led to an enlarged description of familial MDS/AML cases, and the number of mutations involved, suggesting they are more frequent than those previously recognized. Despite the recent discovery of well-established causative gene mutations (RUNX1, GATA2, ETV6, TERT, TERC, SRP72, ANKRD26, DDX41, CEBPA), many cases remain unexplained (about 80%), suggesting that other inherited mutations could predispose to MDS/AML. It is expected that new sequencing approaches will help to the identification of more cases, more genes as well as novel syndromes. In 2017, we started a multicentric prospective study (Clinical trial.gov NCT03058588) aiming to look for predisposing mutations in patients and relatives affected by Familial AML and MDS syndromes (FAMS) by NGS and to screen for old and new mutations potentially associated with the disease. Methods At present, 12 AML/MDS patients have been enrolled. Leukemic (bone marrow) and germline (buccal swab) DNA were analyzed by NGS gene panel approach based on a 28 genes associated to myeloid leukemogenesis, including the 9 above mentioned genes associated to FAMS. NGS libraries were performed by a Nimblegen (Roche) custom panel based on gene capture strategy and the sequencing was performed by MiSeq (Illumina). Results Ten patients did not reveal any germline mutations and the candidates are undergoing to whole exome sequencing. One presented a germline mutation on RUNX1, and the analysis of the affected relatives is on going. One revealed a new mutation. She was a 70 years old woman affected by RARS and her pedigree was characterized by 9 relatives affected by hematologic and solid neoplasia and trombocytopenia (fig 1). The NGS analysis revealed the mutation c.*514C〉T in 3'UTR of ETV6 with VAF of 50% on tumor DNA. The variant has never been described before, while ETV6 has been already associated with FAMS. Sanger sequencing confirmed the mutation on the germline DNA in heterozygosis. The screening of 2 affected relatives still alive confirmed the presence of the variant in heterozygosis. In silico analysis performed on PolymiRST Database revealed that c.*514C〉T in 3'UTR of ETV6 results in a gain of miRNA binding site: hsa-miR- 4717-3p and hsa-miR- 942-3p. Discussion The variant c.*514C〉T in 3'UTR of ETV6 seems to repress ETV6 due to RNA interference. The new binding miRNAs have been already described as over-expressed in solid and hematologic tumors. Moreover, the down-regulation of ETV6 is associated with alteration of cell growth and hematopoiesis. Due to these evidences, c.*514C〉T in 3'UTR of ETV6 could be considered as a new mutation involved in FAMS predisposition. Disclosures No relevant conflicts of interest to declare.

Description: Background Chronic Myeloid Leukemia (CML) is a clonal myeloproliferative disorder which molecular base is represented by the bcr-abl fusion gene, encoding for the constitutionally activated BCR-ABL tirosine-kinase. Three Tirosin-Kinase Inhibitors (TKI) are approved for first line treatment: Imatinib (IM) and the second generation (2G) TKI Nilotinib (NIL) and Dasatinib. 2G TKI are known to provide faster and deeper molecular responses (MR) compared to Imatinib, but serious toxicities may hamper long term treatment with these molecules. Furthermore, 2G TKI were usually employed as second line after IM failure, while the inverse sequence from second to first generation TKI (like an induction-maintenance model) has not been explored yet. We used this schedule in a small group of patients in the PhilosoPhi34 study (EudraCT: 2012-005062-34), a clinical trial designed by the REL (Rete Ematologica Lombarda) cooperative group. This study was composed by three consecutive phases: a Recruitment Phase, a Core Phase (CP) in which patients received NIL 300 mg BID for 12 month (mos), and an Observational Phase (OP), restricted for patients who obtained at least complete cytogenetic response at the end of the CP. During OP, treatment choice was up to the physician and any TKI approved for first line treatment could be used, including IM. In 2017 we presented preliminary data showing that a 12-mos-NIL treatment followed by IM appears as a safe and effective choice for first line therapy in chronic phase CML. Fluctuations in BCR/ABL ratio were similar between IM and NIL treated pts, and the probability of loss of MR4 or MR3 was the same in the two groups; furthermore, despite fluctuations, MR was maintained or improved over time in IM subgroup. Our purpose is to verify these data after 24 mos follow up (FU) at the end of OP. Methods We analyze PhilosoPhi34 database; MR is reported at 3, 6 and 12 mos during the CP and every 6 mos during the OP. The last pt completed the 24 mos of OP in June 2018. Database is still open, evaluations ongoing, and some data can be missing yet: our preliminary observations concern pts with available data of 24 mos OP. Results Seventy-nine pts started the OP. Fourteen pts switched to IM during the OP (Table 1) due to high cardiovascular risk or grade 1-2 chronic AEs . Only 11 pts started IM since the beginning of OP, and we consider these pts in our analysis. Sokal score was high in 2 pts (18%), intermediate in 5 (45.5%), low in 4 (36.5%). At the beginning of OP, 6 pts had a MR ≥ 4 (54.5%), 5 had MR3 (45.5%). At 12 mos of the OP, 7 had MR ≥ 4, 3 had MR3 and 1 had lost MR3 with PCR 0.192%IS (1/5, 20%). At 24 mos of the OP, 9 had MR ≥ 4 (81,8%), and 2 had MR3. Notably, none of pts lost MMR; 2/3 pts(66%) improved response from MR 3 to MR 4 and the pt who transiently lost MMR at 12 mos, recovered it at 24. Sixty-four pts maintained 2G TKI: 62 NIL, 2 other TKI (not considered for analysis). Of them, 4 were lost during this phase: 2 within the first year of OP, other 2 within 12 and 24 mos of OP. In the NIL group, Sokal score was high in 10 pts (16.6%), intermediate in 19 (31.6%) and low in 31 (51.6%). At the beginning of OP, 32 pts had MR ≥ 4 (51.6%), 21 had MR3 (33.8%) and 9 less than MR 3 (14.5%). Responses were improved over time: at 12 mos, 36 pts had MR ≥ 4 (60%), 20 had MR3 (33%) and 4 less than MR3 (6%). At 24 mos 46 pts had MR ≥ 4 (78%), 8 MR3 (13.5%) and 4 less than MR3 (8,5%), Among them, 1 pt experienced disease progression due to a mutation. In particular, during the second year of OP, 11 pts improved response from MR3 to MR ≥ 4(11/20, 55%). Discussion Our data show progressive MR improvement in both IM and NIL group. In particular, risk of loss of MMR is not increased in IM group. More data, more balanced groups and a longer FU are necessary to further confirmations, but after three years of FU, we consider this combination of NIL-followed-by-IM a possible strategy for first line treatment in chronic phase CML, in particular for pts with cardiovascular risk factors. Disclosures Rossi: Janssen: Membership on an entity's Board of Directors or advisory committees; Jazz: Membership on an entity's Board of Directors or advisory committees; Amgen: Membership on an entity's Board of Directors or advisory committees; Sanofi: Membership on an entity's Board of Directors or advisory committees; Roche: Membership on an entity's Board of Directors or advisory committees; Novartis: Honoraria; Gilead: Membership on an entity's Board of Directors or advisory committees; Teva: Membership on an entity's Board of Directors or advisory committees; Mundipharma: Honoraria; BMS: Honoraria; Sandoz: Honoraria; Pfizer: Membership on an entity's Board of Directors or advisory committees; Abbvie: Membership on an entity's Board of Directors or advisory committees; Celgene: Membership on an entity's Board of Directors or advisory committees.

Description: Therapy with Tyrosine Kinase Inhibitors (TKIs) changed the fate of Philadelphia-positive Chronic Myeloid Leukemia (CML). At present, the therapeutic strategy aims to improve the management of the disease and the quality of life of the patients. In July 2015, we started a prospective multicentric randomized trial with the aim to validate the policy of the intermittent de-escalation treatment and to explore the impact of this strategy on the Quality of Life. To this purpose, CML patients older than 60 years in stable (≥2 years) MR3.0 or MR4.0 molecular response were randomized to receive a FIXED intermittent TKIs regimen (one month ON and one month OFF), as previously published (Russo D, Blood 2013; Russo D, BCJ 2015), versus a PROGRESSIVE intermittent TKIs regimen (one month ON and one month OFF for the 1st year; one month ON and two months OFF for the 2nd year; one month ON and three months OFF from the 3rd year) (OPTkIMA study, ClinicalTrials.gov: NCT02326311). Molecular monitoring was performed according to the 2015 ELN guidelines, every 3 months by RT-PCR on peripheral blood (Baccarani M,Blood 2013). In case of MR3.0 (MMR) loss, checked in two monthly consecutive RT-PCR analysis, patients were planned to exit the study and to resume TKIs daily. This first interim report have been focused on the patients who, by intention to treat, have completed the first year of the study for an historical comparison with the previous INTERIM trial (Russo D, Blood 2013; Russo D, BCJ 2015). During the first year, both the patients randomized in the FIXED and in the PROGRESSIVE arms were given the TKIs treatment one month ON and one month OFF. Up to June 2018, 177 patients have been enrolled by 26 Italian Hematological Centers (first patient randomized in July 2015) and 121/177 patients (68%) completed the first year of OPTkIMA study. The median age was 71 years (range 60-89) and 64% of the patients were belonging to the Sokal intermediate/high risk goup. 96/121 (79%), 14/121 (12%) and 11/121 (9%) patients were receiving imatinib (IMA), nilotinib (NILO) and dasatinib (DAS), at the time of enrollment. Overall, 59/121 (49%) and 62/121 (51%) patients have been randomized in the FIXED and PROGRESSIVE arm, respectively. 41/62 patients (66%) randomly assigned to the PROGRESSIVE arm have entered the second year of therapy. 34/121 patients (28%) went out of the study during the first year. The reasons for protocol discontinuation were: informed consent withdrawn (2 cases), second cancer (4 cases), loss of MR3.0 (28 cases). (Table 1). Among the 28 patients who lost the MMR, 17 and 11 were in MR4.0 and MR3.0, respectively, when they were enrolled into the study. Thus the probability of loosing the MR3.0 while on OPTkIMA was 21,7% at one year (Figure 1). All the 28 patients resumed TKIs continuously and all obtained at least the MR3.0 response, within 6 months and are currently included in the study follow up. The intermittent treatment was well tolerated, with 4 serious adverse events (1 appendicitis, 1 atrial fibrillation, 1 cardiac failure, 1 hip fracture) and 3 adverse events (1 diarrohea, 1 pruritus and 1 fever), none of which have been considered treatement-related. None of the patients experienced the TKI withdrawn syndrome. According to this first interim report, we found that a policy of intermittent TKIs administration in elderly patients is safe and well tolerated. Analysis of patient-reported QoL outcomes is ongoing and will further add information on the overall treatment effectivness of the new PROGRESSIVE intermittent TKI administration. After the 1st year, 28/121 patients (23%) lost MR3.0 and all of them re-gained the major molecular response within 6 months from resumption of continuous treatment. The probability of MR3.0 loss while on OPTkIMA at 1 year was 21,7% and this is quite comparable with the 20% MR3.0 loss observed in the previous INTERIM trial (Russo D, Blood 2013; Russo D, BCJ 2015). Disclosures Efficace: Bristol Meyers Squibb: Consultancy; Seattle Genetics: Consultancy; Lundbeck: Research Funding; TEVA: Research Funding; AMGEN: Research Funding; Incyte: Consultancy; Amgen: Consultancy; TEVA: Consultancy; Orsenix: Consultancy. Abruzzese:Novartis: Consultancy; BMS: Consultancy; Pfizer: Consultancy; Ariad: Consultancy. Bonifacio:Incyte: Consultancy; Pfizer: Consultancy; Amgen: Consultancy; Novartis: Research Funding; Bristol Myers Squibb: Consultancy. Castagnetti:Incyte: Consultancy, Honoraria; Pfizer: Consultancy, Honoraria; Bristol Meyers Squibb: Consultancy, Honoraria; Novartis: Consultancy, Honoraria. Giai:Novartis: Consultancy; Pfizer: Consultancy.

Description: AMLs are clonal disorders characterized by high genomic heterogeneity and several chromosomal and molecular alterations affecting patients' outcome. In about 40% of AML patients who do not show any citogenetic alteration, sequencing analysis identified different gene mutations which play a pivotal role in leukemogenesis and have a negative prognostic impact: FLT3, ASXL1, TET2, IDH1, IDH2, RUNX1, CBL, CEBPα, DNMT3A and TP53. Conventional Sanger sequencing may detect clones representing more than 20% of the total tumor population, whereas Next Generation Sequencing (NGS) can identify mutations in less than 1% of leukemic cell burden. The detection of these variants is relevant because they can play an important role in driving drug resistance and disease relapse and for biologic risk assessment. To that purpose, we designed a 23-genes panel including: FLT3, DNMT3A, RUNX1, ASXL1, IDH1, IDH2, BCOR, NRAS, KRAS, TET2, TP53, U2AF1, ZRSR2, SF3B1, SRSF2, CBL, CEBPα, EZH2, NPM1, TERT, TERC, ETV6, GATA2. By using a 454 GS Junior by Roche Diagnostics with an amplicon-based sequencing approach and performing three sequencing runs per sample in order to reach a sensitivity of at least 1%, we settled an investigative strategy in order to assess the genomic profile of AMLs at diagnosis and possibly at the time of relapse or resistance. By this approach, we were able to confirm all the variants (i.e. FLT3, NPM1) previously documented with conventional tests; to reveal the presence of variants under the Sanger threshold of 20% in the genes resulted as wild type with routinely analysis; to evaluate the AML clonal heterogeneity, by assessing the coexistence of several mutated clones which have been described to be related to drug resistance, poor prognosis or to prior myelodysplastic syndrome. This gene panel NGS-based strategy may be proposed as a highly accurate and sensitive approach for genomic characterization of acute myeloid leukemias and as an useful tool for planning a target and personalized AML therapy. Disclosures No relevant conflicts of interest to declare.

Description: Background: Cancer-associated thrombosis (CAT) is a leading cause of death in cancer patients after cancer itself. Risk factors for CAT include tumor type/stage, body mass index (BMI), blood cell counts and chemotherapy exposure. These factors form the basis of prediction algorithms for CAT risk, including most notably the Khorana Risk Score. However, significant limitations exist with these currently-available risk prediction models. Emerging data suggest that a tumor's molecular profile can impact venous thromboembolism (VTE) risk. Mutations of ALK, EGFR, IDH1, ROS1, and KRAS for example have been shown to modulate the risk of CAT; however, these studies were limited by the number of mutations and specific tumor types analyzed. We hypothesized that extended molecular testing in a large patient cohort would allow for improved detection of molecular signatures associated with CAT. We analyzed deep-coverage targeted sequencing data (up to 341 genes) of tumor samples from 11,776 cancer patients to identify gene mutations associated with VTE. Methods: Adult patients with any solid tumor diagnosis who had their tumors sequenced using MSK-IMPACT from 1/2014 to 12/2016 were retrospectively assessed for CAT events using redundant algorithmic methods and individual chart reviews. The endpoint was defined as the first instance of cancer-associated pulmonary embolism and/or proximal/distal lower extremity deep vein thrombosis (DVT). An episode of upper extremity DVT was considered a competing event. The observation period was limited to 365 days after IMPACT blood control sampling. Cause-specific Cox proportional hazards regression was used to test for an association between gene and CAT risk, adjusting for clinical covariates including age, cancer type, cytotoxic chemotherapy (time-dependent), anticoagulant use, stage (metastatic/non-metastatic) and prior history of VTE. Separate multivariate models evaluated the association for the 60 most frequently-mutated genes identified, along with ALK, MET, ROS1 which were included based on existing literature suggesting an effect on VTE risk. Final p-values were adjusted for false discovery using the Benjamini-Hochberg procedure, and the threshold for statistical significance was set at 0.10. Patients with multiple cancer diagnoses were excluded. Results: Out of 11,776 individuals we observed 727 CAT events (6.2% of cohort). The most commonly represented tumor types were lung (18%), breast (15%) and colorectal cancer (10%); see Figure for a breakdown of CAT incidence by tumor type. Most (72%) of patients were metastatic at time of IMPACT testing and 4% were on anticoagulation therapy. Statistically significant predictors of CAT included cytotoxic chemotherapy (HR 1.61 [1.37-1.9]; p