Berry, Lynne D.

INTRODUCTION: Early neoplastic progression of Barrett's esophagus (BE) is often treated with endoscopic therapy. Although effective, some patients are refractory to therapy or recur after apparent eradication of the BE. The goal of this study was to determine whether genomic alterations within the treated BE may be associated with persistent or recurrent disease. METHODS: We performed DNA sequencing on pre-treatment esophageal samples from 45 patients who were successfully treated by endoscopic therapy and did not recur as well as pre-treatment and post-treatment samples from 40 patients who had persistent neoplasia and 21 patients who had recurrent neoplasia. The genomic alterations were compared between groups. RESULTS: The genomic landscape was similar between all groups. Patients with persistent disease were more likely to have pre-treatment alterations involving the receptor tyrosine kinase pathway ( P = 0.01), amplifications of oncogenes ( P = 0.01), and deletions of tumor suppressor genes ( P = 0.02). These associations were no longer significant after adjusting for patient age and BE length. More than half of patients with persistent (52.5%) or recurrent (57.2%) disease showed pre-treatment and post-treatment samples that shared at least 50% of their driver mutations. DISCUSSION: Pre-treatment samples were genomically similar between those who responded to endoscopic therapy and those who had persistent or recurrent disease, suggesting there is not a strong genomic component to treatment response. Although it was expected to find shared driver mutations in pre-treatment and post-treatment samples in patients with persistent disease, the finding that an equal number of patients with recurrent disease also showed this relation suggests that many recurrences represent undetected minimal residual disease.

Abstract Background: Therapies for myelofibrosis (MF) are limited and most are palliative. The JAK1/2 inhibitor ruxolitinib reduces spleen size and MF-related symptoms and improves survival, but can be limited by dose-dependent anemia and thrombocytopenia. Moreover, nearly half of ruxolitinib responders relapse within 5 years. PI3Kd is highly expressed in MF patient samples, independent of ruxolitinib pre-exposure. In JAK2-mutated cell lines, inhibition of PI3K/AKT signaling reduced proliferation and clonogenic potential. The once daily, next generation PI3Kd inhibitor TGR-1202 inhibited PI3K/AKT signaling and led to apoptosis in leukemia and lymphoma cell lines and was well-tolerated in clinical studies, with a toxicity profile distinct from that of ruxolitinib and other PI3Kd inhibitors. We hypothesized that adding TGR-1202 to ruxolitinib could resensitize or augment the response of MF patients with lost or suboptimal response to single-agent ruxolitinib. Objective: To assess safety of TGR-1202 in combination with ruxolitinib in MF patients Secondary Objectives: Hematologic response, symptom assessment Methods: MF patients who had sub-optimal responses to ruxolitinib continued their highest tolerated dose of ruxolitinib without change for ≥ 8 weeks, and were assigned to escalating doses of TGR-1202 in a standard 3+3 algorithm. Adverse events (AEs) were graded by NCI-CTCAE v4.03. Efficacy was assessed according to IWG-MRT consensus response criteria. Symptoms were assessed by the MPN symptom assessment form total symptom score (TSS). All patients received Pneumocystis pneumonia prophylaxis after cycle 1. Results: Eleven MF patients were enrolled and received 400 mg (n=3), 800 mg (n=6), or 600 mg TGR-1202 (n=2) daily. Nine were evaluable for response. Median age was 66y, 73% were male. All had ECOG PS 0-1. Five patients had mutations in JAK2, 4 in CALR, and 3 in MPL; these were mutually exclusive with exception of 1 patient with CALR and MPL mutations (Table 1). Median number of cycles of TGR-1202 + ruxolitinib treatment was 5 (1-13). Grade 2 anemia was the most common AE (Table 2). Two patients had asymptomatic Grade 3 elevations in amylase and lipase that persisted after drug was held, meeting criteria for dose limiting toxicities (DLTs) in 2 separate cohorts (TGR-1202 800mg+ruxolitinib 15mg BID and TGR-1202 800mg+ruxolitinib 10mg BID). Both patients had peak plasma TGR-1202 concentrations 1.5-2x higher than the other patients receiving 800mg TGR-1202, although steady-state levels were equivalent. The maximum tolerated dose (MTD) of TGR-1202 in combination with ruxolitinib was not established. Two patients went off-study due to AEs, and 3 due to progressive disease. One of 9 evaluable patients achieved complete remission and 7 had stable disease. Seven of the 9 evaluable patients had improvement in hematologic parameters and 8 had reduced MF symptoms with a median 33% decrease in TSS (Fig. 1). Conclusions: TGR-1202 + ruxolitinib was well-tolerated. Pharmacokinetic data were consistent with single-agent TGR-1202 (unpublished data), indicating that ruxolitinib does not alter absorption or metabolism of TGR-1202. Grade 3 elevations in amylase and lipase were considered DLTs, per protocol. Although the clinical significance of these asymptomatic laboratory findings is not clear, the protocol was amended to further assay these labs and to exclude concomitant medications with the potential to increase amylase/lipase. Importantly, no grade ≥3 hepatotoxicity, colitis, or thrombocytopenia was seen and no MTD was found. Although only one patient achieved CR, 89% demonstrated clinical benefit with the addition of TGR-1202 to ruxolitinib, supporting further exploration of this combination. Disclosures Strickland: Alexion Pharmaceuticals: Consultancy; Ambit: Consultancy; Baxalta: Consultancy; Boehringer Ingelheim: Consultancy, Research Funding; CTI Biopharma: Consultancy; Daiichi Sankyo: Consultancy; Sunesis Pharmaceuticals: Consultancy, Research Funding; Abbvie: Research Funding; Astellas Pharma: Research Funding; Celator: Research Funding; Cyclacel: Research Funding; GlaxoSmithKline: Research Funding; Karyopharm Therapeutica: Research Funding; Sanofi: Research Funding. Miskin:TG Therapeutics, Inc: Employment, Equity Ownership. Cavers:TG Therapeutics: Employment, Equity Ownership. Sportelli:TG Therapeutics, Inc.: Employment, Equity Ownership. Michaelis:Pfizer: Equity Ownership; Cellgene Corporation: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Incyte Corporation: Consultancy, Honoraria. Mesa:CTI: Research Funding; Promedior: Research Funding; Celgene: Research Funding; Gilead: Research Funding; Incyte: Research Funding; Galena: Consultancy; Ariad: Consultancy; Novartis: Consultancy. Savona:Amgen Inc.: Membership on an entity's Board of Directors or advisory committees; TG Therapeutics: Research Funding; Ariad: Membership on an entity's Board of Directors or advisory committees; Takeda: Research Funding; Sunesis: Research Funding; Incyte: Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene: Membership on an entity's Board of Directors or advisory committees; Gilead: Membership on an entity's Board of Directors or advisory committees.

Purpose Indications for sentinel lymph node biopsy (SLNB) for thin melanoma are continually evolving. We present a large multi-institutional study to determine factors predictive of sentinel lymph node (SLN) metastasis in thin melanoma. Patients and Methods Retrospective review of the Sentinel Lymph Node Working Group database from 1994 to 2012 identified 1,250 patients who had an SLNB and thin melanomas (≤ 1 mm). Clinicopathologic characteristics were correlated with SLN status and outcome. Results SLN metastases were detected in 65 (5.2%) of 1,250 patients. On univariable analysis, rates of Breslow thickness ≥ 0.75 mm, Clark level ≥ IV, ulceration, and absence of regression differed significantly between positive and negative SLN groups (all P < .05). These four variables and mitotic rate were used in multivariable analysis, which demonstrated that Breslow thickness ≥ 0.75 mm (P = .03), Clark level ≥ IV (P = .05), and ulceration (P = .01) significantly predicted SLN metastasis with 6.3%, 7.0%, and 11.6% of the patients with these respective characteristics having SLN disease. Melanomas < 0.75 mm had positive SLN rates of < 5% regardless of Clark level and ulceration status. Median follow-up was 2.6 years. Melanoma-specific survival was significantly worse for patients with positive versus negative SLNs (P = .001). Conclusion Breslow thickness ≥ 0.75 mm, Clark level ≥ IV, and ulceration significantly predict SLN disease in thin melanoma. Most SLN metastases (86.2%) occur in melanomas ≥ 0.75 mm, with 6.3% of these patients having SLN disease, whereas in melanomas < 0.75 mm, SLN metastasis rates are < 5%. By using a 5% metastasis risk threshold, SLNB is indicated for melanomas ≥ 0.75 mm, but further study is needed to define indications for SLNB in melanomas < 0.75 mm.