Search

Immune checkpoint blockade has revolutionised the treatment of multiple cancers including melanoma, non-small cell lung cancer, urothelial and renal cell cancers. For patients with chemorefractory gastroesophageal cancer, treatment with anti-PD-1 therapy results in modest benefits in overall survival; nivolumab and pembrolizumab have been licenced in Japan and the USA, respectively, for this indication. However, initial enthusiasm has been tempered by the results of several large negative trials; immune checkpoint blockade is not superior to chemotherapy in the second-line setting or beyond in unselected or low PD-L1-expressing patients. Microsatellite instability is uncommon in patients with metastatic gastric cancer; however, it is associated with response rates of more than 50% and long-term survival benefit. Combining anti-PD-1 with cytotoxic chemotherapy and targeted therapies also shows promise to extend the benefit of immune checkpoint blockade to a larger proportion of gastroesophageal cancer patients. In this review we discuss recently reported and ongoing clinical research in immunotherapy for gastroesophageal cancer, and consider molecular biology associated with sensitivity and resistance to immune checkpoint blockade in gastroesophageal cancer patients.

Background Loss of ataxia telangiectasia mutated (ATM), a key protein regulating DNA repair signaling, has been suggested to increase sensitivity to DNA damaging agents. We conducted a study analyzing the loss of ATM protein expression in colorectal cancer and correlated this with clinical outcomes. Materials and Methods The clinical outcomes data and tumor samples from metastatic colorectal cancer patients referred to the Royal Marsden Hospital Drug Development Unit (United Kingdom) from 2012 to 2016 and providing consent for a molecular characterization study were analyzed. Immunohistochemistry (IHC) slides were assessed by a pathologist for nuclear staining intensity of ATM and semiquantitatively scored. ATM loss was defined as a nuclear H-score of ≤ 10. Results Of 223 colorectal cancer samples, ATM IHC loss was identified in 17 (8%). ATM loss was independent of the RAS and RAF mutational status. ATM loss was associated with superior overall survival after first-line oxaliplatin-based therapy (49 vs. 32 months; hazard ratio [HR], 2.52) but not with irinotecan-based therapy (24 vs. 33 months; HR, 0.72). ATM loss was not prognostic for survival from the diagnosis (50 vs. 44 months; HR, 1.43). Conclusion ATM could be considered a biomarker for the development of novel DNA repair targeting agents and treatment of colorectal cancer.

Background The incidence and clinical significance of electrolyte abnormalities (EAs) in phase I clinical trials are unknown. The objective of this study is to evaluate the incidence and severity of EAs, graded according to CTCAE, v4.03, to identify variables associated with EAs and their prognostic significance in a phase I population. Methods A retrospective chart review was performed of 1088 cases in 82 phase I clinical trials consecutively treated from 2011 to 2015 at the Drug Development Unit of the Royal Marsden Hospital. Cox regression analysis was performed to examine the relationship between overall survival (OS) and baseline characteristics, treating the occurrence of grade III/IV EAs as a time-varying covariate. Results The most common emergent EAs (all grades) were as follows: hyponatraemia 62%, hypokalaemia 40%, hypophosphataemia 32%, hypomagnesaemia 17% and hypocalcaemia 12%. Grade III/IV EAs occurred in 19% of cases. Grade III/IV EAs occurred during the dose-limiting toxicity window in 8.46% of cases. Diarrhoea was associated with hypomagnesaemia at all grades (p, Highlights • Patients who develop grade III/IV electrolyte abnormalities have poorer overall survival. • Hyponatraemia is linked to higher risk of developing other electrolyte abnormalities. • Hypoalbuminaemia is linked to higher risk of developing electrolyte abnormalities.

Gastroesophageal adenocarcinoma (GEA) has long been one of the most common cancer worldwide, yet its underlying molecular abnormalities were only recently described. Advances in genomics, epigenetics, transcriptomics and proteomics, amongst others, allowed an unprecedented insight into the molecular mechanisms driving gastric and esophageal cancer carcinogenesis. Spearheaded by the Cancer Genome Atlas project, molecular classification of GEA revealed different subtypes, based on microsatellite instability, Epstein-Barr virus positivity, somatic copy-number variation and chromosomal instability, whilst other groups provided classifications based different characteristics, namely specific gene expression signatures, impaired DNA repairs pathways and mutational load. Microsatellite instability remains one of the most important biomarkers in GEA, as it is associated with a better prognosis and significant responses to immuno-oncology agents. Current therapeutics strategies for GEA include targeting receptor tyrosine kinase amplifications (HER-2, EGFR or c-MET), DNA repair pathways (BRCA1/2 mutations, ATM-loss), angiogenesis (VEGR, VEGFR2) or immune checkpoint inhibition (PD-1/-L1 axis, CTLA-4). Novel targets in GEA were described, with promising clinical results, and involve the tumoural microenvironment (matrix metalloproteinase), tight gap junction (Claudin 18.2) and the WNT developmental pathway (DKK-1). Potential futures approaches include targeting epigenetics modification, for example micro-RNAs, DNA methylation and histone acetylation. In conclusion, recent advances in oncogenomics enabled the identification of a multitude of therapeutic targets. As such, solid biomarker work that integrate all the available data is needed to guide treatment selection; that is undoubtably what will be of most benefit for GEA patients

Item does not contain fulltext BACKGROUND: Adolescent and young adult (AYA) patients with advanced solid tumours are often considered for phase I clinical trials with novel agents. The outcome of AYAs in these trials have not been described before. AIM: To study the outcome of AYA patients in phase I clinical trials. METHODS: Clinical trial data of AYAs (defined as aged 15-39 years at diagnosis) treated at the Drug Development Unit, Royal Marsden Hospital, between 2002 and 2016, were analysed. RESULTS: From a prospectively maintained database of 2631 patients treated in phase I trials, 219 AYA patients (8%) were identified. Major tumour types included gynaecological cancer (25%) and sarcoma (18%). Twenty-five (11%) had a known hereditary cancer syndrome (most commonly BRCA). Molecular characterisation of tumours (n = 45) identified mutations most commonly in TP53 (33%), PI3KCA (18%) and KRAS (9%). Therapeutic targets of trials included DNA damage repair (16%), phosphoinositide 3-kinase (PI3K) (16%) and angiogenesis (16%). Grade 3/4 toxicities were experienced in 26% of patients. Of the 214 evaluable patients, objective response rate was 12%, with clinical benefit rate at 6 months of 22%. Median overall survival (OS) was 7.5 months (95% confidence interval: 6.3-9.5), and 2-year OS was 11%. Of patients with responses, 36% were matched to phase I trials based on germline or somatic genetic aberrations. CONCLUSION: We describe the outcome of the largest cohort of AYA patients treated in phase I trials. A subgroup of these patients demonstrates benefit, with several durable responses beyond 2 years. A sizeable proportion of AYA patients have cancer syndromes, significant family history or somatic molecular aberrancies which may influence novel therapeutic treatment options.

The intestinal microbiota plays a key role in the pathogenesis of acute graft-versus-host disease (aGVHD). High-dose conditioning regimens given prior to allogeneic hematopoietic stem cell transplantation (aHSCT) modulate the composition of gut microbiota and damage the gut epithelial barrier, resulting in increased systemic inflammation. We assessed whether gut decontamination with antibiotics (ATB) prior to aHSCT influenced the frequency of aGVHD and mortality in 500 patients from two Canadian centers between 2005 and 2012. The rate of grade II–IV aGVHD was higher in the ATB arm compared with the arm without ATB (42% vs 28%; p < 0.001). This difference was mainly driven by a 2-fold higher rate of grade II–IV gastrointestinal aGVHD (GI-GVHD) in the ATB arm compared with the arm without ATB (20.7% vs 10.8%; p = 0.003). Multivariate analyses adjusted for known aGVHD risk factors revealed that more patients in the ATB group developed clinically significant GI-GVHD and liver aGVHD; adjusted odds ratio (aOR) = 1.83; p = 0.023 and aOR = 3.56; p = 0.047, respectively. Importantly, median overall survival (OS) was significantly lower in the group receiving ATB and the OS at 10 y remained decreased in the ATB group; adjusted hazard ratio (aHR) = 1.61 (p < 0.001).

Introduction Small cell lung cancer (SCLC) is aggressive and relapse is inevitable. Novel therapies in the context of phase I (Ph1) clinical trials should be considered in this setting. Next generation sequencing (NGS) identifying putative driver mutations may aid therapy allocation in such trial settings. Methods Retrospective analysis of characteristics and clinical outcomes of patients (pts) with advanced SCLC referred to the Phase I Clinical Trials Drug Development Unit at the Royal Marsden Hospital between 1992 and 2016, using electronic patient records Results 85 pts with advanced SCLC were included, of which 45 pts were allocated to ≥1 trial each (48 allocations). Of these, 21 (46.7%) pts started ≥1 trial, with a total of 24 trials commenced. Of the 64 pts who were unable to commence trials, 43 (67.2%) had rapid disease progression, 11 (17.2%) sought other therapies, 5 (7.8%) declined Ph1 trials, and 5 pts did not enrol for other reasons. Pts had a median of 2 previous lines of treatment (range:1-3). 13 pts had tumor molecular profiling with high coverage targeted NGS, yielding 22 distinct mutations; Aberrations in DNA damage repair (DDR) pathway were most common: TP53 (53.8%), FANC (30.8%), ATM (15.4%), ATR, BARD1 and CHEK2 (7.7% each). All pts found to have ≥1 DDR pathway mutation had platinum sensitive disease (disease progression ≥90 days from last platinum dose) in first line setting. Other mutations found include ALK, APC, AR, AXIN1, DDB2, ERBB2, mTOR, PI3K, NOTCH, NTRK17, PDGFR, RB1, RET, STK11, FKT3 and VGFR2. Pts were most commonly treated with novel inhibitors against PARP (n=9), PI3K/AKT (n=3), PD-1 (n=2) and BCL (n=2). Therapies were generally well tolerated with no dose limiting toxicities observed. Of 9 pts who received PARP inhibitors, 1 achieved a RECIST partial response (PR) and remained on trial for 16 weeks; 4 achieved RECIST stable disease (SD) for a median of 17 weeks (range:10.9-31.1 weeks). 1 exceptional responder whose tumor harbored multiple somatic aberrations (TP53, FANCF, AR, ERBB2, NOTCH2) was enrolled on 3 sequential phase I trials - she achieved durable RECIST SD on a PARP inhibitor (31.1 weeks), RECIST PR with a PD-1 inhibitor (38.7 weeks), and durable RECIST SD with carboplatin/ATR inhibitor (27.9 weeks). Platinum sensitivity in the first line setting predicted for improved disease control (RECIST PR/SD 73.3% vs 0%, p Discussion Novel therapeutic agents within the context of a dedicated Ph1 trials unit were well tolerated and led to preliminary signals of antitumor activity. Targeted NGS may aid in the identification of pts with putative gene aberrations, potentially predicting for response to novel therapies. Ph1 trials should be considered earlier in the pt's disease course to maximize their prospects of trial enrolment. Citation Format: Joline S. Lim, Samuel J. Harris, Malaka Ameratunga, Raghav Sundar, Joo Ern Ang, Dearbhaile Collins, Maxime Chénard-Poirier, Alvaro I. Garces, Stan B. Kaye, Juanita Lopez, Udai Banerji, Johann S. de Bono, Timothy A. Yap. Precision medicine for patients with advanced small cell lung cancer treated with novel therapeutic agents in a phase I clinical trials unit [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 3081. doi:10.1158/1538-7445.AM2017-3081

2542 Background: The use of oral anticancer therapies has increased substantially in recent years. Nonadherence can impair the efficacy of such therapies as well as confound the interpretation of toxicity and pharmacokinetic data in phase I trials. However, there is a paucity of data regarding adherence patterns and barriers in this specific setting. Methods: We included patients treated in Phase I trials involving oral investigational medicinal products (IMPs) in the Drug Development Unit, Royal Marsden Hospital, UK, between 2012 and 2014. Patient, disease and treatment characteristics as well as compliance data from prospectively collected trial diary cards and drug accountability were recorded. Relationships between adherence rate and other variables were explored using logistic regression. Results: We collected data for 2819 patient-weeks, pertaining to 169 patients treated on 18 different phase I trials. Median age was 61 years (range 18-79), females predominated (60%), median number of previous systemic therapy was 3 (0-12) and median time on trial was 9 weeks (0.3-212.4). Hundred-percent adherence rate was 88% in the first cycle and 79% overall. Nonadherence occurred in 83 of 2819 patient-weeks (3.0%); including 75 (2.7 %) missed doses and 8 (0.3 %) overdoses. In univariate analysis, longer time on trial and a continuous treatment schedule were associated with poorer adherence, whereas fasting requirements pre- or post-dosing was associated with improved adherence. Known intracranial metastases, number of concomitant medications and opiates or anti-emetics use were not significantly associated to adherence. In multivariate analysis, fasting requirements (OR 5.347, 95%CI : 1.443-20.019, p = 0.012) and longer time on trial (OR 0.98, 95%IC : 0.961-0.996, p = 0.017) were statistically significant. Conclusions: This is the first report on adherence rates to oral anticancer IMPs in a large phase I trial population. Our observed adherence rates are at the higher end of published data in the general cancer population. Factors influencing adherence in phase I trials appear to be distinct, with fasting requirements being a unique finding. These findings may impact future early-phase trial design and conduct.

2506 Background: RO5126766 is a potent RAF and MEK inhibitor with activity in xenografts models of RAS and RAF-mutated cancers. We present data from the RAS/RAF-mutated advanced solid tumor cohort and the initial results for the multiple myeloma (MM) cohort. Methods: Patients with KRAS, NRAS or BRAF-mutant tumours were treated with RO5127566 using a novel schedule:4mg twice weekly in 4-week cycles. For MM patients, it was given 3 weeks out of 4 and co-administration of weekly dexamethasone was authorised. Response assessment was completed using RECIST 1.1 criteria for solid tumours and the International Myeloma Working Group (IMWG) criteria were used for MM. Results: A total of 20 patients with solid tumours (10 NSCLC, 5 gynaecological cancers and 5 miscellaneous cancers) and 1 MM patients were evaluable. Among the 10 KRAS-mutant NSCLC patients, tumour regression was seen in 6/10 (60 %), of which 3/10 (30 %) were partial responses. Two of these patients had maintained response for over 1 year and one patient is still on study after 30 cycles. Of the gynaecological cancers, 3/5 patients (60%) achieved a partial response ( KRAS-mutant endometrial and ovarian cancer and BRAF-mutant ovarian). Of these patients, 1 of the KRAS mutants had received 2 previous lines of MEK inhibitors and the BRAF mutant had previously received a BRAF inhibitor. In the miscellaneous group, 4 patients with colorectal cancer (2 BRAF and 2 NRAS) and 1 patient with NRAS-mutant melanoma were treated and none responded. Two patients with MM have been treated so far (1 KRAS, 1 KRAS+NRAS). The one evaluable patient has had an IMWG partial response (PR) after 1 cycle (FLC-λ from 324 mg/L to 161mg/L, ratio 0.03 to 0.08) without concomittant dexamethasone. This patient was previously treated with an immunomodulatory drug, a proteasome inhibitor and two ASCTs. Conclusions: RO5126766 has shown exciting preliminary activity across a wide range of RAS- and RAF-mutated malignancies, with significant response rates in lung and gynaecological cancers. To our knowledge, the PR seen in our MM patient represents one of the first responses to a single-agent RAF/MEK inhibitor in multiple myeloma in a trial context. Clinical trial information: NCT02407509.

TPS2607 Background: SRA737 is a highly selective, orally bioavailable small molecule inhibitor of Checkpoint kinase 1 (Chk1), a key cell cycle checkpoint and central regulator of the DNA Damage Response (DDR) network. In cancer cells, replication stress induced by genomic alterations in oncogenes (eg, MYC and RAS) combined with loss of function in tumor suppressors (eg, TP53 and ATM) results in persistent DNA damage and genomic instability. Targeted inhibition of components of the DDR network such as Chk1 by SRA737 may be synthetically lethal to cancer cells and have utility as a monotherapy in a range of tumor indications. SRA737 is currently being investigated in two Phase 1 trials in patients with advanced cancer. We now describe the Phase 1 multicenter, dose-escalation, monotherapy study of SRA737 (NCT02797964). Methods: Up to 40 patients with advanced cancer will receive oral SRA737 administered daily on a 28-day schedule. For dose-escalation, an accelerated titration design with 100% dose escalation and single patient cohorts is allowed until Grade 2 related toxicity is observed, followed by a rolling-6 design. Dose expansion will include 6 patients with any solid tumor treated at the recommended Phase 2 dose (RP2D). Eligibility criteria include WHO performance status of 0-1 and ≤ 3 prior lines of cytotoxic chemotherapy for metastatic disease. Primary objectives are to assess the safety profile of monotherapy SRA737 and to establish a RP2D. The PK profile and PD biomarkers will be investigated. The study was opened to enrollment in mid-2016. An amendment, which includes the addition of indication specific expansion cohorts of subjects with genetically-defined tumors known to have Chk1-sensitizing aberrations such as gene mutations and amplifications/deletions, has been submitted and is pending regulatory review while enrollment continues. At the Annual Meeting, the amended design will be described. The dose and schedule identified in this trial will inform the design and conduct of Phase 2 studies of SRA737 as a single agent and in combination with other targeted or immunomodulatory agents. Clinical trial information: NCT02797964.

TPS2613 Background: SRA737 is a highly selective, orally bioavailable small molecule inhibitor of Checkpoint kinase 1 (Chk1), a key cell cycle checkpoint and central regulator of the DNA Damage Response (DDR) network. In cancer cells, replication stress induced by genomic alterations in oncogenes (eg, MYC and RAS) combined with loss of function in tumor suppressors (eg, TP53 and ATM) results in persistent DNA damage and genomic instability. Targeted inhibition of components of the DDR network such as Chk1 by SRA737 may be synthetically lethal to cancer cells. Chk1 is also believed to facilitate tumor cell resistance to chemotherapy or radiation-induced DNA damage and the combination of SRA737 with these standards-of-care may provide synergistic antitumor activity. SRA737 is being investigated in two Phase 1 trials in patients with advanced cancer. We now describe the Phase 1 multicenter, dose-escalation study of SRA737 in combination with gemcitabine/cisplatin (GC) or gemcitabine (G) alone (NCT02797977). Methods: Up to 70 patients will receive escalating doses of SRA737+GC in Stage 1 or SRA737+G in Stage 2 until a recommended Phase 2 dose (RP2D) is established, followed by expansion cohorts. Patients will receive a single SRA737 PK run-in dose followed by Gem on D1 and 8, Cis on D1, SRA737 on D2, 3, 9 and 10 of each 21-d cycle or Gem on D1, 8 and 15, SRA737 on D2, 3, 9, 10, 16, 17 of each 28-d cycle. Eligibility criteria include WHO performance status of 0-1 and ≤ 3 prior lines of cytotoxic chemotherapy for metastatic disease. Primary objectives are to assess the safety profile of SRA737 combination therapy and to establish a RP2D. The PK profile and PD biomarkers (eg, phosphorylation of Chk1 at Ser296, Ser345 and γH2AX foci in PBMCs and tumor tissue) will be explored. The study was opened to enrollment in mid-2016. An amendment, which includes the addition of an indication specific expansion cohort of subjects with genetically-defined tumors known to have Chk1-sensitizing aberrations (eg, gene mutations and amplifications/deletions), has been submitted and is pending regulatory review while enrollment continues. At the Annual Meeting, the amended design will be described. Clinical trial information: NCT02797977.

10536 Background: AYA cancer patients are relatively under-represented in clinical trials, with no published data regarding their outcomes in phase I studies. Trials utilizing novel therapeutic agents are often considered in these patients, due to their tendency to have good organ reserve, and ability to tolerate additional lines of therapy. This study describes the experience of AYA patients with advanced solid tumors treated in a specialized drug development unit. Methods: Patient characteristics and clinical outcomes of AYA patients (defined as age 15 to 39 years at time of initial cancer diagnosis) treated at the Drug Development Unit, Royal Marsden Hospital, United Kingdom, between 2002 and 2016, were captured and analyzed from case and trial records. Results: From a database of 2631 patients treated on phase I trials, 219 AYA patients (8%) were identified. Major tumor types included gynaecological cancer (24%), sarcoma (18%), gastrointestinal (16%) and breast cancer (11%). Patients had a median of 3 previous lines of systemic chemotherapy (range 0 – 6), and 19% participated in 2 or more phase I studies. Twenty (9%) had a known hereditary cancer syndrome (most commonly BRCA), 27% had a family history (FH) of cancer, 15% no FH and 49% no FH documented. Molecular characterization of tumors (n = 45) identified mutations most commonly in p53 (33%) , PI3KCA (18%) and KRAS (9%) . Major trial categories included DNA damage repair (16%), PI3K (16%) and anti-angiogenesis (15%) agents. Grade 3/4 toxicities were experienced in 25% of patients (10% hematological). Of the 214 evaluable patients, objective response rate was 12%, with clinical benefit rate at 6 months of 22%. Median progression free survival was 2.3 months (95% CI: 1.9 to 2.8), median OS was 7.6 months (95% CI: 6.3 to 9.5), and 2-year OS was 11%. Of patients with responses, 35% were matched to phase I trials based on germline or somatic genetic aberrations. Conclusions: A sub-group of AYA patients with advanced solid tumors derive considerable benefit from participating in trials involving novel therapeutics. Future research must focus on predictive biomarkers and molecular profiling to identify those that would benefit from novel therapies.

8556 Background: Malignant pleural mesothelioma (MPM) has been historically documented as a locally infiltrative disease in large series from the early 1980s. With the changing landscape of cancer diagnosis and treatment, increased areas of unusual metastases have been published as case reports. With no standard second-line therapies for MPM, referral to early phase trial units is common. We report the metastatic patterns of a large cohort of MPM patients treated at the Royal Marsden Drug Development Unit (DDU). Methods: Clinical data was gathered for MPM patients referred to the DDU from 1992 to 2016. Radiographic details were collected from CT, bone scan and FDG PET imaging. Prior treatment, response, somatic mutations, clinical trial and survival data was obtained from medical records. Results: From the database, 165 evaluable patients with MPM were identified. Median age at diagnosis was 64 years (range 37–90) and 76% were male. Epithelioid MPM comprised 81% and 65% were right sided. Bone metastases were reported in 20%, with the majority lytic in nature ( Table). Peritoneal and omental disease was evident in 24% with ascites in 16%. In 11% of cases lung metastases presented as diffuse miliary-type pattern. Visceral metastases (15%) were predominantly liver (78%), but also occurred in adrenals, spleen and kidneys. Symptomatic brain metastases were recorded in 3%. Median overall survival was 24.2 months (95% CI: 20.8 - 29.2). Conclusions: This large study documents the metastatic patterns of advanced MPM in the 21st century and highlights an increased frequency of traditionally unexpected sites of metastases. Higher than expected incidence of lytic bone metastases (20%) suggests consideration of bone imaging in advanced MPM clinical workflow and trial protocols. [Table: see text]

Introduction: The impact of commensal bacteria harbored in the gastrointestinal tract known as the microbiota has long been recognized as a pivotal factor in allogeneic hematopoietic stem cell transplantation (aHSCT). The microbiota is at the origin of acute graft-versus-host disease (aGVHD) and infections, two important lethal complications in aHSCT. High-dose conditioning chemotherapy prior aHSCT disrupts the gut epithelial barrier allowing bacterial by-products to translocate into the peripheral blood and modulates T cell response and pro-inflammatory cytokines. Such observation led to an effort by certain transplant centers to eliminate bacterial colonization prior to aHSCT to decrease the risk of gram-negative bacterial translocation. In this study, we assessed whether patients' gut decontamination prior to allogeneic stem cell infusion influenced the frequency of aGVHD, pneumomatosis coli (PC) a significant complication of gastrointestinal (GI) GVHD and overall survival. Methods: We retrospectively reviewed the charts of 543 patients who had undergone a single myeloablative or nonmyeloabaltive aHSCT for hematological malignancies from two academic hospitals in the province of Quebec, Canada between January 2005 and December 2012. Exclusion criteria included prior aHSCT, syngeneic and haploidentical aHSCT. Each university hospital has implemented a different pre-transplant antibioprophylaxis guideline. At HMR hospital, ciprofloxacin or moxifloxacine were started at initiation of the conditioning regimen for gut decontamination, but were omitted in patients with fluoroquinolone or penicillin allergy and during nosocomial infections outbreak. On the other hand, in the CHU de Quebec patients were not prescribed prophylactic antibiotics (ATB). In addition, ATB used to treat infections before the stem cells infusion were considered. To determine the impact of ATB, we performed multivariable analyses adjusting for the following confounding factors: age, gender, stem cell origin (source), donor type/match, and conditioning regimens to compare the frequency of aGVHD, PC, leucocytes recovery at day+14 and overall survival (OS) in patients receiving or not ATB. Results: 500 patients were included and a total of 240 (48%) patients received ATB at the time of conditioning regimen. Demographics were similar in both groups with mean age of 48 years. Frequency of grade II-IV aGVHD was more elevated in patients receiving ATB compared to the no ATB group (42% vs 28% respectively with adjusted OR (aOR)=1.53 (p Discussion: This retrospective study indicated that ATB were associated to a more severe aGVHD driven by digestive manifestations of the GVHD and higher incidence of PC even after multivariable analysis. These life-threatening complications may impact on the 1-year and OS that were lower in the ATB group. Without undermining the role of ATB prophylaxis to prevent infection in aHSCT setting, treatment with ATB impact on the commensal microbiota and diminishes its diversity. This imbalance created by the ATB may have contributed to the pathophysiology of GI-GVHD. This ultimately highlights the importance to reconsider antibioprophylaxis to preserve an intact flora and its benefits in aHSCT. Disclosures No relevant conflicts of interest to declare.