Median Overall Survival For Pts Research Articles

Abstract PO5-06-08: Real-world analysis of efficacy, adverse events and predictive biomarkers for advanced triple negative breast cancer (TNBC) treated with immune check point Inhibitors (ICI): A single center experience

Abstract Introduction KEYNOTE-355 investigated the addition of pembrolizumab to chemotherapy in advanced TNBC patients (pts) with PD-L1 positive tumors showing median overall survival (OS) of 23.0 months (mos) and median progression-free survival (PFS) of 16.1 mos. Given significant OS benefit with pembrolizumab, it is the current standard of care while the indication of atezolizumab has been withdrawn due to lack of OS benefit in IMPASSION-130. Clinical efficacy and adverse events in pts on this treatment regimen in clinical practice is unknown and understanding the real-world outcomes of this regimen is critical. Methods We conducted a retrospective, single-center observational study among TNBC pts treated with ICI (Pembrolizumab or Atezolizumab) at Roswell Park Comprehensive Cancer Center from January 2017 to May 2023. Demographics and clinicopathological variables were collected including comorbidities, laboratory data, sites of metastases (mets), treatment received, immune related adverse events (irAEs), and clinical outcomes. Adverse events were reported using the Common Terminology Criteria for Adverse Events v5.0. Serial CT scans were reviewed and response rate was determined using RECIST v1.1. Patient demographic, clinical and outcome characteristics were summarized by treatment and survival outcomes were obtained using standard Kaplan-Meier methods. Associations between survival outcomes and baseline or treatment characteristics were evaluated using Cox regression models using Firth’s method. All analyses were conducted in SAS v9.4 at a significance level of 0.05. Results A total of 44 pts with advanced TNBC treated with ICI were included (23 received pembrolizumab and 21 received Atezolizumab). The study population consisted of all female pts with stage IV disease, median age 53.7 years (IQR 30.4-85.2), 68.2 % (30/44) White, 25.0% (11/44) Black, 34.1% (15/44) obese (BMI≥30). 27.3% (12/44) pts developed any grade irAEs which included myocarditis (4.5%), rash (9.1%), hyperthyroidism (6.8%), hypothyroidism (4.5%), adrenal insufficiency (2.3%), diabetes mellitus (2.4%), colitis (2.3%), and transaminitis (2.3%). Distribution of reported irAE was 25% grade 1 (3/12), 66.7% grade 2 (8/12) and 8.3% grade 3 (1/12). 75% pts (33/44) received standard treatment (ICI+ chemotherapy) of which 72.7 % (24/33) received it in the first-line setting. Median OS in pts treated in first line was 16.2 mos (95% CI, 10.7-NR) and median PFS was 4.4 mos (95% CI, 2.7-8.4). Overall response rate (ORR) was 29.1% (7/24) of which 8.3% (2/24) had complete response and 20.8% (5/24) had partial response. ORR was 22.2 % (2/9) in pembrolizumab cohort and 33.3 % (5/15) in atezolizumab cohort. Among pts with first line treatment, obese pts were found to have improved PFS compared to non-obese (11.5 vs 3.8 mos, univariate hazard ratio (HR) 0.46, 95% CI 0.23-0.93, p= 0.031), and this difference was maintained in multivariable analysis even after adjusting for age (adjusted HR (aHR) 0.40, 95% CI 0.17-0.98, p= 0.044). Pts with brain mets had poor extracranial PFS compared to those without brain mets (2.8 vs. 5.2 mos, HR 2.25, 95% CI 1.08-4.68, p= 0.036), however this difference was not observed when further adjusted for age (aHR 2.27, 95% CI= 0.91-5.68, p= 0.08). Conclusion Clinical outcomes in our study were inferior to KEYNOTE-355 where median PFS was 16.1 mos and OS was 23.0 mos and ORR was 52.7% for PD-L1 positive population. This may reflect a more heterogeneous population of pts treated in routine clinical practice who are typically less fit than pts on clinical trials. Our study found improved outcomes among obese patients, similar to data reported in other disease settings. These data warrant multi-center validation with larger number of patients. Citation Format: Archit Patel, Arya Mariam Roy, Malak Alharbi, Kristopher Attwood, Chi-Chen Hong, Song Yao, Thaer Khoury, Amy Early, Tracey O'Connor, Ellis Levine, Shipra Gandhi. Real-world analysis of efficacy, adverse events and predictive biomarkers for advanced triple negative breast cancer (TNBC) treated with immune check point Inhibitors (ICI): A single center experience [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO5-06-08.

Tagraxofusp in Combination with Azacitidine and Venetoclax in Newly Diagnosed CD123+ Acute Myeloid Leukemia, Expansion Cohort of a Phase 1b Multicenter Trial

Background: CD123 is a subunit of the interleukin 3 (IL3) receptor expressed on the surface of blasts in most acute myeloid leukemias (AMLs). Tagraxofusp (TAG), a CD123-directed therapy, is a recombinant protein drug consisting of IL3 fused to a truncated diphtheria toxin payload approved for treatment of blastic plasmacytoid dendritic cell neoplasm. We found TAG resistance in AML cells was mediated by DNA methylation and silencing of diphthamide genes (e.g., DPH1), which causes resistance to diphtheria toxin (Togami, JCI 2019). The hypomethylating agent azacitidine (AZA) reversed TAG resistance, and TAG plus AZA prolonged survival vs. either alone in xenograft models. TAG-exposed AML cells were also sensitized to the BCL2 inhibitor venetoclax (VEN). We performed a phase 1b trial combining these agents in myeloid malignancies and reported acceptable safety in dose escalation of TAG with AZA-VEN (ASH 2021). Here, we report an expansion cohort in newly diagnosed AML treated with TAG-AZA-VEN. Methods: Patients (pts) were adults with AML ineligible or declining intensive induction chemotherapy. CD123+ blasts were required (tested locally by flow or IHC, any level). Eligibility included albumin >3.2 g/dL and normal cardiac ejection fraction. Pts were treated on 28-day cycles with AZA 75 mg/m2 d1-7, VEN 400 mg d1-21 with ramp up in cycle 1, and TAG 12 µg/kg d4-6 ( Fig A), with cycle 1 hospitalization to mitigate risks of capillary leak syndrome (CLS). Also included are 3 pts who received TAG 7 µg/kg (n=2) or 9 µg/kg (n=1) in escalation. TAG was given on d4 to avoid VEN ramp up overlap and because preclinical data showed AZA pretreatment can sensitize AML cells to TAG. Results: 26 pts with AML were treated; median age 71 (range 60-81). All were adverse risk per ELN 2022 criteria, including 13 (50%) with a mutation in TP53 and others with 1 or more of ASXL1, BCOR, EZH2, RUNX1, SF3B1, SRSF2, STAG2, or U2AF1. Eight (31%) had complex karyotype, 8 (31%) secondary AML, and 5 (19%) therapy-related AML. Adverse events (AEs) were like those seen with TAG or AZA/VEN, without evidence of exacerbation by combination. Grade 3+ AEs occurring in >15% of pts regardless of attribution included thrombocytopenia (54%), leukopenia (54%), febrile neutropenia (35%), and anemia (31%). CLS, a known TAG side effect, occurred in 19% (n=5; three grade 2, one gr 3, one gr 4; all in cycle 1) and was manageable by albumin supplementation and diuresis. Mortality at 30 days was 11.5%, with causes of death of sepsis (n=2), multiorgan system failure (n=1). 18 of 26 pts (69%) achieved a best response of CR (10; 39%), CRi (5; 19%), or MLFS (3; 12%). Of the 13 with TP53 mutation, 7 (54%) achieved CR/CRi/MLFS (CR=4, CRi=2, or MLFS=1). Bone marrow blasts decreased in all pts who had a biopsy after starting treatment ( Fig A). For the 18 achieving CR/CRi/MLFS, the median duration of response was 12.4 months (mo) (95% CI, 6.1-NA). 13 of 26 (50%) proceeded to allogeneic stem cell transplantation, which included 6/13 (46%) with TP53 mutation. The median number of cycles was 3 for both pts who did (range 1-15) and did not (range 2-4) proceed to transplant. The median follow-up was 10.7 mo, including post-transplant timepoints. Median overall survival (OS) was 14 mo (95% CI, 9.5-NA) and progression-free survival (PFS) was 8.5 mo (95% CI, 5.1-NA). In pts with TP53 mutation, median OS was 9.5 mo (95% CI, 1.8-NA); and not reached (NR) without ( Fig B). PFS for pts with TP53 mutation was 5.1 mo (95% CI, 1.8-NA), and 13.3 mo (95% CI, 8.6-NA) without. Measurable residual disease (MRD) was assessed by flow cytometry in 17 pts who achieved CR/CRi/MLFS; 12/17 (71%) were MRD negative (<0.1%). 4/7 (57%) with TP53 mutation achieving CR/CRi/MLFS were MRD negative. Median OS for pts who became MRD negative was NR. Median OS for the 13 pts who received allogeneic transplant was 18.2 mo (95% CI, 14.0-NA). Median PFS for the transplanted pts was 13.3 mo (95% CI, 8.2-NA). Conclusions: Treatment of AML with triplet TAG-AZA-VEN is feasible and shows encouraging activity as up-front therapy in adverse risk AML, including in pts with TP53 mutation. These data (OS 14 mo; TP53 mutant OS 9.5 mo) compare favorably to HMA/VEN in ELN adverse risk (OS 12 mo; Lachowiez, BloodAdv 2023) or TP53 mutant AML with poor risk cytogenetics (OS 5.2 mo; Pollyea, CCR 2022), albeit with some differences between populations. No new safety signals were observed. Further development of this combination is warranted, including directly compared to AZA-VEN.

Neighborhood Socioeconomic Disadvantage and Distance from Treatment Center Do Not Impact Survival Outcomes of Patients with Non-Hodgkin Lymphoma and Multiple Myeloma Treated with CAR T-Cell Therapies

Introduction: Chimeric antigen receptor (CAR) T-cell therapies have radically changed the treatment paradigm of patients (pts.) with relapsed/refractory (r/r) hematologic malignancies. Despite their remarkable efficacy, health care disparities can impact outcomes of pts. in the real-world setting. Living in a disadvantaged neighborhood has been associated with inferior outcomes after allogenic hematopoietic stem cell transplantation (HSCT), independent from individual-level socioeconomic (SE) factors. Our study aimed to identify the impact of neighborhood adversity on survival outcomes of recipients of CAR T-cell therapies, as this has not been yet described. Methods: We retrospectively identified consecutive adult pts. with r/r B-cell non-Hodgkin lymphoma (B-NHL) and multiple myeloma (MM) who received CAR T-cell therapies at our center from May 2018- January 2023. Neighborhood disadvantage was defined by Area Deprivation Index (ADI), a validated tool that allows for ranking of census block groups based on SE domains (income, education, employment, and housing quality) at a national level. The 2021 national ADI percentile ranks (1-100) were obtained for each pt. using the 9-digit zip code of place of residence, with a higher rank corresponding to higher level of disadvantage. We compared overall survival (OS) and progression free survival (PFS) between pts. with high vs. low ADI. We also examined survival outcomes in relation to estimated distance from place of residence to treatment center (DTC). Time dependent outcomes were calculated from date of CAR T infusion. Pts. with B-NHL and MM were analyzed separately. Baseline variables were compared between groups using chi-squared, t-test, and Wilcoxon rank sum tests as appropriate. Survival estimates were calculated with Kaplan-Meier method, and hazard ratios were generated using cox proportional hazards model. Results: CAR T therapy was administered in 124 and 45 pts. with B-NHL and MM, respectively. Median ages were 65 and 62 yrs., 68% and 51% were male, 94% and 84% were White, 47% and 20% had high-risk disease (based on IPI/MIPI for B-NHL, and FISH for MM), 30% and 69% had prior HSCT, and median number of prior therapies were 3 (range 1-6) and 6 (range 4-14), among pts. with B-NHL and MM, respectively. B-NHL subtypes included DLBCL (84%), MCL (6%), FL (5%), and PMBCL (5%), and CAR T products for B-NHL were axi-cel (52%), tisa-cel (24%), liso-cel (18%) and brexu-cel (6%). Ide-cel (82%) and cilta-cel (18%) were used in pts. with MM. We treated pts. from 169 census block groups over 8 states, although 89% pts. were from Ohio. For the entire cohort, median ADI rank was 62.5 (range 1-100) and median DTC was 42.5 (range 1-4559) miles (m). Pts. were categorized (based on median) as having high (> 62.5) or low (≤ 62.5) ADI, and long (> 42.5m) or short (≤ 42.5m) DTC. Longer DTC was significantly associated with higher ADI ( p<0.001). Baseline characteristics (age, sex, prior therapies, prior HSCT, high risk disease, CAR T product) were similar across groups with low vs. high ADI and short vs. long DTC. Median follow up time up was 12 (range 4-60) and 9 (range 4-22) months for living pts. with B-NHL and MM, respectively. For pts. with B-NHL, objective response rates (ORR) (85% vs. 80%; p=0.4), relapse rates (RR) (66% vs. 66%; p=0.9), median OS (19 vs. 14 months; HR of death: 1.2; 95% CI 0.7-1.8; p=0.5) ( Fig.1), and median PFS (10 vs. 5 months, HR of relapse: 1.1, 95% CI 0.7-1.7; p=0.6) were similar for pts. with low vs. high ADI. For pts. with MM, there were no differences in ORR (77% vs. 87%; p=0.3), RR (77% vs. 70%; p=0.5), median OS (14 vs. 18 months; HR of death: 0.8; 95% CI: 0.3-2; p=0.6) ( Fig.2), and median PFS (7 vs. 9 months, HR of relapse: 0.7, 95% CI 0.3-1.4; p=0.3) between those with low vs. high ADI. There were no differences in median OS of pts. with short vs. long DTC for NHL (19 vs. 14 months; HR: 1.2, 95% CI 0.7-1.9, p=0.4) or MM (14 vs. 20 months; HR: 0.9, 95% CI 0.3-2.5; p=0.9), or when DTC was analyzed as a continuous variable. Conclusion: In patients with r/r B-NHL and MM who received CAR T-cell therapies, response rates and survival outcomes were comparable regardless of neighborhood disadvantage level or distance to treatment center. Pts. living in more disadvantaged neighborhoods travelled longer distances for treatment. These findings prompt future investigation into referral patterns and access barriers to cellular immunotherapy, especially in disadvantaged neighborhoods.

The effect of antibiotic (Abx) use on immune checkpoint inhibitor (ICI) efficacy in patients (pts) with metastatic urothelial carcinoma (mUC) in a real-world setting.

4580 Background: Use of Abx may negatively affect outcomes with ICI, possibly due to negative impact on gut microbiota. We previously showed that in our real world cohort of 138 mUC pts, Abx use within 30 or 60 days (d) of ICI initiation did not impact overall survival (OS). We now report updated follow-up (fu) and results from our expanded cohort of 310 mUC pts who received ICI. Methods: We performed a retrospective analysis of adult pts with mUC treated at the Cleveland Clinic between 2015 and 2022 who received 2 or more cycles of ICI with pembrolizumab (P), atezolizumab (A) or avelumab (Av). Effect of Abx use within 30 and 60 (d) of ICI initiation was correlated with progression free survival (PFS) and OS. OS and PFS were estimated using Kaplan-Meier method and compared using log rank test. Results: Median age of pts was 71 years (41-90) and 75% pts were males. 224 pts (72.3%) received P, 82 (26.4%) A and 4 (1.3%) Av. Fluoroquinolones (31.1%), penicillins (29.6%) and cephalosporins (19.7%) were most commonly used Abx. With a median fu of 47 weeks (3.86- 325.3), median OS in pts who received Abx within 60 (d) prior to ICI was 11.79 months (mos) vs 14.55 mos in those who did not receive Abx (p=0.007) and median PFS was 3.68 mos vs 5.29 mos respectively (p=0.038). Median OS in pts who received Abx within 30 days prior ICI was 10.41 mos vs 14.46 mos in those who did not receive Abx (p=0.037). There was a non-significant trend for worse PFS in pts who received Abx within 30 (d) prior to ICI (3.06 mos vs 5.03 mos, p=0.238). There was no effect on OS or PFS with Abx use 30 or 60 (d) after ICI initiation. Conclusions: In our real-world cohort of mUC pts receiving ICI, Abx use 60 and 30 (d)prior to ICI initiation was associated with significant worse OS and worse PFS for the former group but Abx use after ICI initiation did not have an effect on OS or PFS. These findings have the potential to influence clinical practice, including a higher threshold for giving Abx to mUC pts starting ICI. Further prospective studies are warranted. [Table: see text]

Demographic and clinical characteristics of patients with metastatic breast cancer and leptomeningeal disease: A single center retrospective cohort study.

1108 Background: Patients (pts) with metastatic breast cancer (MBC) and leptomeningeal disease (LMD) have a poor prognosis with limited therapeutic options. It is critical to better understand the risk factors and natural history of this condition, including pts in a modern cohort. Methods: In this single center retrospective cohort study, we identified pts with radiographic and/or cytologic evidence of LMD who received care at the University of California San Francisco (UCSF) from 2000-2023. To identify pts, we used the UCSF Pathology Database, the UCSF Radiation Oncology Database, and EMERSE, a searchable copy of the medical record. We conducted chart review to identify demographic and clinical characteristics, treatment history, and overall survival (OS). Results: 106 pts with MBC and radiographic and/or cytologic evidence of LMD were identified. All but one pt was female (n=105; 99.1%); most pts were non-Hispanic (n=93, 87.7%) and white (n=74, 69.8%). Median age at the time of MBC diagnosis was 51.9 years; 25 pts (20%) had de novo metastatic disease. Most pts had invasive ductal carcinoma (n=71, 66.9%), with invasive lobular carcinoma making up 21.7% (n=23) of cases. Pts presented with the following BC subtypes: Hormone receptor (HR) positive, HER2 negative (n=50, 47.1%), HR+/HER2+ (n=19, 17.9%), HR-/HER2+ (n=9, 8.5%), and triple negative (TNBC; n=23, 21.7%). Most pts had brain metastasis (n=78, 73.6%) in addition to LMD. Median time from diagnosis of MBC to LMD was 17.6 months (0-101.2 months). Pts received a median of 3.0 lines of therapy prior to the diagnosis of LMD (range 0-11.0) and 1.0 lines after the diagnosis of LMD (range 0-4.0). Many pts received intrathecal (IT) therapy (n=42, 39.6%) and/or whole brain radiation therapy or spinal radiation (n=49, 46.2%). Median OS from the diagnosis of LMD to death was 3.3 months. There was no significant difference in median OS by subtype: HR+/HER2- 3.5 months, HER2+ 4.7 months, and TNBC 2.2 months (p=0.45). There was no significant difference in median OS by histology: ductal (3.3 months) vs. lobular (4.7 months) (p=0.56). Pts who received IT therapy survived longer than those not treated with IT therapy (4.7 vs. 2.5 months, p=0.01). There was no significant difference in median OS among pts diagnosed with LMD between 2000-2015 (2.9 months) vs. 2016-2023 (3.6 months) (p=0.74). Conclusions: This study represents one of the largest reported case series of pts with MBC and LMD, including a more modern cohort. There appeared to be an over-representation of pts with lobular cancer and those with de novo metastatic disease. Most pts had brain metastases in addition to LMD. Survival was short in all pts, particularly those who did not receive IT therapy. There was no significant difference in median OS in pts diagnosed with LMD from 2000-2015 vs. 2016-2023; this remains an area of unmet clinical need.

HER3 expression status following systemic chemotherapy treatment in biliary tract cancers.

e16153 Background: Erb-b2 receptor tyrosine kinase 3 (HER3) is frequently overexpressed in a variety of cancers. Activation of HER3 signaling serves an important role in promoting cell proliferation and is associated with chemoresistance. Thus, there is interest in exploring HER3 status of tumors as a potential prognostic and/or therapeutic target. Prior studies have reported HER3 status in various cancers after treatment with chemotherapy (CT). However, HER3 expression profile in biliary tract cancer (BTC) remains unknown. Here, we evaluated HER3 status of patients with BTC post CT. Methods: The medical records of patients (pts) who had been initially diagnosed with BTC and treated at the National Cancer Center Hospital (Tokyo, Japan), between January 2010 and June 2020, were retrospectively analyzed. Pts whose post CT specimens were available were included in this study. Immunohistochemical staining for HER3 was performed using HER3/ErbB3 (D22C5) XP Rabbit mAb (Cell Signaling Technology) as a primary antibody. IHC scoring (0, 1+, 2+, 3+) of membranous staining intensity was performed according to HER2 IHC gastric scoring guideline. Results: A total of 19 pts with BTC, with a median age of 59 (range 44-79) years, had specimens available after post CT and were evaluated for HER3 expression status. HER3 status was 3+ in 10 pts, 2+ in 3 pts, 1+ in 4 pts, and 0 in 2 pts. The specimens were obtained from liver in 17 pts, bone in 1 pt, and omental lymph node in 1 pt. The histological type was adenocarcinoma in all pts, and HER3 status according to primary tumor type is shown in the table below. Fourteen pts were treated with gemcitabine (GEM) plus cisplatin (CDDP), 1 pt with GEM plus CDDP plus S-1, 1 pt with GEM plus S-1 as first-line CT for advanced BTC, and 3 pts with S-1 as an adjuvant treatment. Fifteen pts were biopsied after CT to participate in clinical trial. The remaining 4 obtained specimens after CT during primary surgery or surgery for metastases or for cholangitis. The median overall survival of pts with advanced BTC was 23.6 months in HER3 3+, 17.1 months in HER3 2+, 27.7 months in HER3 1+, and 24.6 months in HER3 0. Four pts had specimens available from pre and post CT, all of which were intrahepatic cholangiocarcinoma with the treatment of GEM plus CDDP. HER3 status was upregulated in 3 pts (1 pt 1+ to 2+, 2 pts 2+ to 3+) and downregulated in 1 pt (1+ to 0) at post CT. Conclusions: HER3 expression (1+, 2+, 3+) was observed in 89.5% of previously treated pts with BTC, and HER3 status was upregulated after CT in some pts. These results suggest that HER3 might be a potential therapeutic target in BTC post CT. [Table: see text]

Survival outcomes in duodenal carcinoma based on facility type: A National Cancer Database analysis.

e16328 Background: Duodenal Carcinoma (DC) is a rare entity, accounting for 0.5% of gastrointestinal tumors and up to 50% of all small intestinal tumors. Few studies have characterized survival outcomes in DC due to its rarity. This study aims to evaluate if Academic centers (AC) have better survival outcomes compared to Non-academic centers (NAC). Methods: A retrospective analysis of the National Cancer Database was conducted from 2004-2019 for subjects with DC, identified by ICD code C17.0 (n=26,070). Features were compared between patients (pts) at AC (n=15,020) and NAC (n=11,050). Overall survival (OS) was estimated using Kaplan Meier method and difference in OS was measured using pairwise log-rank test. Multivariate analysis was performed using binary logistic regression. Results: The median OS in pts with DC was 44 months. Treatment at AC was associated with a longer median OS than NAC (65 vs 32 months, p<0.001). Pts with higher Charleson/Deyo scores (CDS) of 2-3 were less likely to go to AC (OR 0.75, 95% CI 0.67-0.83, p<0.001). AC attracted age ranges <65 years compared to >70 years (OR 0.81, 95% CI 0.74 - 0.88, p<0.001) and >85 years (OR 0.53, 95% CI 0.47-0.60, p<0.001). African Americans (AA) and Asians were more likely to go to AC compared to Whites [(AA OR 1.21, 95% CI 1.13-1.3, p<0.001); (Asian OR 1.43, 95% CI 1.24-1.65, p< 0.001)]. Spanish/Hispanic pts were more likely to go to AC than non-Spanish/Hispanic (OR 1.26, 95% CI 1.13-1.41, p<0.001). Pts in upper income quartiles were more likely to be treated at AC (OR 1.19, 95% CI 1.08-1.32, p<0.001). Pts with Medicare (OR 0.71, 95% CI 0.60-0.83, p<0.001) were less likely to go to AC compared to uninsured pts. Pts from Central USA (OR 0.67, 95% CI 0.63-0.70, p <0.001) and the West Coast (OR 0.46, 95% CI 0.43-0.70, p<0.001) were less likely to go to AC compared to pts from the East Coast. Pts in rural areas were less likely to go to AC compared to those in metros (OR 0.70, 95% CI 0.57-0.87, p<0.001). The 30-day readmission rates after surgery were also studied. Unplanned 30-day readmission rates for surgically treated pts at AC were higher than those at NAC (OR 1.15, 95% CI 1.00-1.31, p=0.041). Conclusions: The median OS of pts with DC who were treated at AC was significantly longer compared to those treated at NAC. Pts with higher CDS were less likely to go to AC. AC had higher unplanned 30-day readmission rates for surgically treated pts than NAC. Pts with DC treated at AC were more likely to be uninsured, younger, and from ethnic minorities. Significant geographical differences also existed where pts from the East Coast and metros were more likely to go to AC compared to those in the West coast, Central and rural areas. Geography may play a role in serving densely populated urban centers where socio-economic factors drive larger minority population to AC facilities. Nonetheless, the difference in overall survival between AC and NAC deserves additional studies to improve outcomes of pts served by NAC.

241P APOBEC3-induced PIK3CA mutations predict better clinical outcomes with PI3K inhibitors in patients with PIK3CA-mutated advanced breast cancer

APOBEC3 enzymes are a key source of mutagenesis in breast cancer. PI3K inhibitors are currently used in the treatment of PIK3CA-mutated advanced breast cancer. However, whether specific PIK3CA mutations (muts) are associated with better outcomes is unclear. We developed a computational framework to screen for APOBEC3-induced mutations in PIK3CA within the TCW context using whole-exome sequencing (WES) data from the breast cancer cohort of the cancer genome atlas (TCGA-BRCA) and an independent cohort of patients (pts.) with ER+/HER2- PIK3CA-mutated advanced breast cancer who were treated with PI3K and aromatase inhibitors (AI). Using the WES of 696 pts with luminal disease from the TCGA-BRCA cohort, we identified 296 pts (42.5%) with PIK3CA muts. Of these, our screen retrieved 104 pts (35.1%) with APOBEC3-induced PIK3CA muts (A3P). There was no significant difference in the median overall survival in pts with A3P compared to non-A3P mutant pts (129 vs. 114 months, p=0.49), and neither significant difference in TMB (1.3 vs 1.17 mut/Mb, p=0.75). ESR1 protein was significantly less expressed in A3P compared to non-A3P mutant samples (log2 ratio: -0.42, p=0.04). The majority of A3P muts (95%) were found in the helical domain of PIK3CA. A3P muts were numerically higher in lobular vs. ductal subtypes (43% vs. 32%, p=0.08). We validated our prioritized A3P muts using a cohort of 65 pts with PIK3CA-mutated ER+/HER2- advanced cancer. The rate of A3Pm was 45% (n=29) which was higher than the rate in pts with non-metastatic cancer in TCGA-BRCA. Pts with A3P muts were significantly associated with longer median progression-free survival on PI3K inhibitor and AI (5 vs. 2 months, hazard ratio: 0.56, 95%CI:0.31-1.00, p=0.02) and a trend toward higher objective response rate (26% vs. 5%, p=0.06). We identified a patient with the A3P E453K muts who achieved partial response for 11 months on alpelisib and AI. This mutation is not routinely included in the companion diagnostic test. Our findings indicate that using a prioritized catalog of APOBEC3-induced mutations in PIK3CA could identify driver mutations that confer oncogenic addiction, hence better response to PI3K inhibitors.

Abstract PD13-09: PD13-09 Clinical outcomes of patients with HR+ HER2- advanced breast cancer with early progression on CDK4/6 inhibitors

Abstract Background: CDK4/6 inhibitors (CDK4/6i) paired with endocrine therapy (ET) are considered first-line (1L) therapy for patients (pts) with HR+ HER2- advanced breast cancer (aBC). A minority of pts will demonstrate primary resistance to CDK4/6i, as characterized by early progression. Thymidine kinase 1 (TK1) is a cell-cycle regulated enzyme downstream of CDK4/6 and involved in nucleotide metabolism during DNA synthesis. Prior studies have shown TK1 may serve as a biomarker of response to CDK4/6i, with early TK1 activity (TK1a) suppression after initiation of CDK 4/6i therapy associated with improved PFS. Lack of TK1a suppression may be associated with primary resistance to CDK4/6i. In this study, we aim to analyze response to subsequent lines of therapy and overall survival (OS) of pts with early progression on 1L CDK4/6i. Methods: Pts with HR+ HER2- aBC from a phase II trial of an alternative schedule of palbociclib (palbo alt dosing trial NCT 3007979) and from a retrospective palbociclib study were included in this analysis. Pts in the palbo alt dosing trial underwent baseline and C1D15 TK1a analysis after initiation on CDK4/6i. C1D15 TK1a suppression was defined at TK1a < 30 Du/L. Pts in the retrospective palbociclib study included pts receiving palbo as part of their standard of care 1L therapy for HR+ HER2- aBC at Washington University in Saint Louis from 2016 to 2021. Clinical information, including treatment start and stop dates on each of the next-line therapies, were collected from the electronic medical record. PFS was estimated by the treatment duration on a specified treatment regimen. Early progression on CDK4/6i was defined as PFS < 6 mo. Best response was defined as next line of therapy with the numerically longest PFS. OS was defined as time to death from the initiation of CDK4/6i. Results: Of the 54 pts enrolled on the palbo alt dosing trial, 51 pts were evaluable for clinical benefit and 46 pts were evaluable for TK1a suppression rate at C1D15. 7 pts (15.2%) were found without TK1a suppression at C1D15. This lack of TK1a suppression on palbo was associated with a significantly shorter PFS (median PFS=3.1 mo) compared to not reached in pts with TK1a suppression at C1D15. We conducted clinical analysis on N=26 pts who exhibited early progression on CDK4/6i which included 10 pts from the palbo alt dosing trial and 16 from the retrospective study. The average subsequent line of therapies in this cohort was 3, with the most common second line (2L) therapy being chemotherapy (N=17, 65.4%) and ET (N=8, 30.8%). The median PFS for pts receiving 2L chemotherapy and ET was 4.09 mo and 3.64 mo, respectively. 10 pts received both chemotherapy and ET with 7 (70.0%) achieving best response with chemotherapy compared to 3 pts (30.0%) who achieved best response with ET. The median OS for the cohort was 14.6 mo. Conclusions: Early progression on CDK4/6i is associated with a particularly poor prognosis. In our cohort, the median OS was far below the expected median OS for pts receiving 1L palbo as reported in the PALOMA-2 trial (14.6 mo vs 53.9 mo). Early progression on CDK4/6i is associated with more aggressive disease which may respond more favorably to chemotherapy, as demonstrated by best response to therapy. Further prospective studies are warranted to explore this treatment approach. Citation Format: Katherine K. Clifton, Shana N. Thomas, Jingqin Luo, Jing Xi, Nusayba A. Bagegni, Foluso O. Ademuyiwa, Rama Suresh, Ashley Frith, Andrew A. Davis, Ron Bose, Katherine Weilbaecher, Whitney L. Hensing, Timothy Pluard, Massimo Cristofanilli, Hyo S. Han, Adam M. Brufsky, Kevin Kalinsky, Shom Goel, Seth A. Wander, Lindsay L. Peterson, Cynthia Ma. PD13-09 Clinical outcomes of patients with HR+ HER2- advanced breast cancer with early progression on CDK4/6 inhibitors [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD13-09.

Abstract PD7-05: Survival of patients with brain metastases at initial breast cancer diagnosis over the last decade

Abstract Background Brain metastases (BM) is a challenging presentation of breast cancer (BC) and has historically been associated with poor prognosis. The risk of BM is higher in patients (pts) with hormone receptor (HR) negative, human epidermal growth factor receptor 2 (HER2)-positive, or triple negative (TN) tumors. There have been significant advances in the treatment of metastatic BC over the past several years, and long-term outcomes after a diagnosis of breast cancer brain metastases (BCBM) are lacking; therefore, this study aimed to identify clinical predictors of BM at initial BC diagnosis, to describe trends in overall survival (OS) over the past decade, and to identify factors associated with OS after BM diagnosis. Methods We evaluated pts with de novo stage IV BC using the Surveillance, Epidemiology and End Results (SEER) database from 2010-2019. A multivariate logistic regression model was conducted among all pts with stage IV BC to assess predictors of BM at initial BC diagnosis. OS probabilities were estimated using the Kaplan-Meier method and log rank test was used to compare differences between groups. Among pts with BM at initial BC diagnosis, univariate analyses were performed to determine the effects of each variable on OS. A Cox proportional hazards regression was used to assess the independent association of several variables with OS. Results A total of 425,110 pts with BC were identified from 2010-2019; 25,113 had stage IV BC at diagnosis and among these, 1,939 pts had BM at initial BC diagnosis. For this last group median age was 60 years and median follow up was 54 months. We performed a logistic regression model to evaluate the factors correlated with BM at initial BC diagnosis among all stage IV BC pts and found that lobular vs ductal histology (OR 0.68; 95% CI 0.55-0.85), HR-/HER2+ vs HR+/HER2- tumors (OR 1.93; 95% CI 1.64-2.29), and the presence of bone (OR 1.19; 95% CI 1.07-1.32), liver (OR 1.25; 95% CI 1.12-1.39), lung (OR 1.72; 95% CI 1.55-1.89), or lymph node (OR 1.21; 95% CI 1.03-1.42) metastases, were significantly associated with this presentation. The table shows OS by tumor subtype among all pts with BM at initial BC diagnosis, as well as the survival rate at 2, 5 and 8 years. We observed significant differences in survival by tumor subtype, where pts with HR+/HER2+ disease had the longest OS (median 18 months; 95% CI: 13-22 months; p = < 0.0001) and the highest rate of OS at 8 years (12.2%; 95% CI 7.61%-17.95%). In contrast, TN BC had a median OS of 6 months (95% CI 5-8 months) and a rate of OS at 8 years of 2.57% (95% CI 0.1%-5.48%). Multivariate analysis among pts with BM at initial BC diagnosis revealed differences in OS in those who were >64 years vs < 50 years (Hazard Ratio [HzR] 1.51; 95% CI 1.29-1.77), married vs single (HzR 0.79; 95% CI 0.70-0.91), lower vs higher income (HzR 1.45; 95% CI 1.19-1.76), tumor grade III/IV vs grade I (HzR 1.68; 95% CI 1.24-2.28), TN subtype vs HR+/HER2- (HzR 2.13; 95% CI 1.81-2.49), as well as pts with additional liver (HzR 1.34; 95% CI 1.20-1.51) or lung metastases (HzR 1.34; 95% CI 1.20-1.49). We did not observe significant changes in OS over time (adjusted HzR 0.996 per year; 95% CI: 0.96-1.02, p = 0.837). Conclusions: Over the last decade, the median OS of pts with BM at initial BC diagnosis has remained poor. However, a substantial minority of pts survive 5 or more years, with long-term survival rates higher in pts with HER2+ tumors. In addition to tumor subtype, OS varied according to age, the presence of metastases to other organs, and sociodemographic factors. Citation Format: Jorge Avila, Julieta Leone, Carlos T. Vallejo, Nancy U. Lin, Jose P. Leone. Survival of patients with brain metastases at initial breast cancer diagnosis over the last decade [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD7-05.

Long-Term Follow-up of Bendamustine Plus Rituximab Regimen in 69 Treatment Naïve (TN) Patients with Waldenström Macroglobulinemia, a Study on Behalf of the French Innovative Leukemia Organization (FILO)

Background Waldenström macroglobulinemia (WM) is a rare hematological malignancy characterized by the accumulation of IgM-secreting clonal lymphoma cells with a lymphoplasmacytic morphological differentiation. Recently, BTK inhibition has emerged as a promising therapeutic strategy in WM (Buske, JCO 2022; Tam, Blood 2020). However, in real life, a large part of patients(pts) continue to be treated by chemoimmunotherapy (Buske, Lancet Hematol2018) including bendamustine (B)-rituximab (R) in salvage (Tedeschi , Leukemia Lymphoma 2015) or first line MW (Treon, ASH2016). We have previously reported a series of 69 WM pts treated by BR (Laribi, BJH 2018) and demonstrated that this regimen is efficient. However, the follow-up was relatively short and some questions remained unresolved in terms of long-range response and second cancer occurrence. Here we report results of this series with a longer follow-up. Methods Sixty-nine symptomatic previously untreated pts with WM diagnostic criteria according to the second WM Workshop recommendations were enrolled in this retrospective multi-center study between January 2013 and December 2017 in 13 French centers. MYD88L265P and CXCR4 mutations were searched using qPCR and or next generation sequencing. Del(17p) and/or TP53 mutation were investigated in 34 pts. All pts received a BR regimen consisting of 375 mg/m2 of R on day 1 and 90 mg/m2 of B on days 1 and 2, repeated every 4 weeks, with a maximum of 6 cycles. Cumulative incidence of secondary malignancies and high-grade lymphomas were computed, considering death free of malignancies as a competing event. Event-free survival (EFS), time to next treatment (TNT), progression-free survival (PFS) and overall survival (OS) were plotted using the log rank test and Kaplan-Meier graphical comparison. Results Pts median age was 69 years old (yo; range 45-88) at WM diagnosis and 71 yo (range 46-88) at therapy initiation. MYD88 and CXCR4 mutations were present in 45/51 (88%) and 11/44 (25%) of pts, respectively. A complex karyotype was disclosed in 6/44 (14%) pts, and 1/34 (3%) had TP53 alteration. According to the IPSSMW score, 14 (22%) pts were low risk, 32(46%) intermediate, and 32 (46%) high risk. Using consensus response criteria, the overall response rate was 97% (very good partial response [VGPR] with negative immunofixation 19%, VGPR 37%, PR40%, MR 1%) and the major response rate was 96%. Cumulative incidence of VGPR or better was 47.8%, 53.6%, 55%, and 56% at 6, 12, 18, and 24 months, respectively. After a median observation time of 68.5 months (range 23.13-106) post BR, the median OS was not reached (83-NR) , and the median PFS was 82 months (75-NR). Median OS for pts with mutated or WT MYD88 was NR and 72.8 months, respectively (p=0.046). Median OS for pts with mutated or WT CXCR4 was 70.2 months and NR, respectively (p=0.64). Median OS related to disease or treatment was not reached, while median OS related to other causes was 106 months. Sixteen (23.2%) pts relapsed. Fourteen received second line therapy and 2 supportive care. Eight patients received ibrutinib, and 6 chemo-immunotherapy. TNT was 31.55 months (range 2.66-82.43). Median PFS after a second line of treatment (PFS2) was 45 (9.13-NR) or 21 months (range 9.13-NR), for pts receiving ibrutinib or not. Eleven pts developed a second cancer: 9 solid tumors (2 pancreatic, 2 gastric, 1colic, 1 pulmonary, 1 breast, 1 cutaneous), and 2 treatment-related myeloid neoplasms (TRMN). Median EFS, defined as occurrence of death, relapse or second cancer was 82 months (61-NR). The cumulative incidence of second cancer was 2.90%, 5.80%, 10.49%, and 17.6% at 12, 24, 48, and 96 months, respectively (figure 1). Univariate analyses on OS disclosed statistically significant differences for age at diagnosis (p=0.001), age at first therapy (p <0.001), IgM>70g/L (p: 0.018), complex karyotype (p=0.037), and high risk IPSSMW (p<0.01). In multivariate analysis, only IPSSMW remained statistically significant (p<0.001) (table1). Univariate analyses on PFS only singled out age at diagnosis (p=0.03) or at first therapy (p=0.049) as statistically significant. Mutated MYD88 or CXCR4 did not impact OS nor PFS. Conclusion This study demonstrates that the BR regimen is efficient in treatment naïve WM pts, yielding long-term responses. The occurrence of secondary cancers, including TRMN, should be closely monitored in these patients. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Treatment Patterns and Outcomes of Patients with Primary or Secondary Acute Myeloid Leukemia By Type of Site (Academic or Community/Government): A CONNECT® Myeloid Registry Study

Introduction Up to 30% of acute myeloid leukemias (AML) evolve from prior hematologic disorders and are classified as secondary AML. Secondary AML is traditionally associated with worse prognosis than de novo, or primary, AML, resulting in inferior response rates and poorer survival. The treatment landscape for AML has evolved recently with the introduction of venetoclax (VEN) and the availability of targeted agents. To better understand evolution of the management of secondary AML, we evaluated treatment patterns and outcomes in patients (pts) with primary AML vs secondary AML treated in real-world clinical practice and enrolled in the CONNECT® Myeloid Disease Registry (NCT01688011) by type of site (academic or community/government). Methods Pts were classified as having either primary or secondary AML based on history of myelodysplastic syndrome (MDS) or progression from MDS to AML. Overall survival (OS) from the date of AML diagnosis, estimated by the Kaplan-Meier method, was evaluated in each group according to type of site (academic or community/government), utilization of allogeneic hematopoietic stem cell transplantation (alloHSCT), and tumor protein p53 (TP53) mutational status. Time to adoption of a VEN-based therapy, defined as first use of VEN-based therapy following Food and Drug Administration (FDA) approval in this setting, was evaluated using a Cox model with factors for age (< 75 and ≥ 75 y), type of site, and geographic region. Results A total of 891 pts in the Registry with a diagnosis of AML were assessed (primary AML, n = 676; secondary AML, n = 215). Median (range) ages were 70 (55-97) y in the primary AML group and 74 (19-93) y in the secondary AML group. Most pts were male (primary AML, 61.6%; secondary AML, 61.5%) and most were treated at community or government centers (primary AML, 60.2%; secondary AML, 66.0%). Pts with primary AML had significantly longer survival than pts with secondary AML; median OS (95% confidence interval [CI]) was 15.6 (13.6-17.4) mo and 5.1 (3.6-6.7) mo, respectively (P < 0.001). Median OS for pts with primary AML and secondary AML treated at academic centers was 17.0 mo and 5.6 mo, respectively (hazard ratio [HR] [95% CI], 0.41 [0.30-0.54], P < 0.001) and in community/government centers was 14.1 mo and 5.2 mo, respectively (HR [95% CI], 0.53 [0.43-0.66], P < 0.001). There was no significant difference in median OS in pts treated at academic centers compared with pts treated at community/government centers for both primary AML (academic vs community/government centers: 17.0 mo [95% CI, 14.7-20.7] vs 14.1 mo [11.0-16.9]; Figure 1A) and secondary AML (5.0 mo [95% CI, 3.0-7.2] vs 5.1 mo [3.7-7.3]; Figure 1B). Significantly longer median OS in pts with primary AML vs secondary AML was observed regardless of alloHSCT status (no alloHSCT: 11.7 mo vs 4.5 mo; HR [95% CI], 0.53 [0.44-0.64], P < 0.001; with alloHSCT: 64.3 mo [35.8-NC] vs 24.9 mo [6.8-48.9], P = 0.002]. Longer median OS was observed among pts who received alloHSCT compared with those without alloHSCT: primary AML, 41.1 mo vs 12.8 mo; secondary AML, 24.9 mo vs 4.4 mo; additional analyses are warranted to account for immortal time bias. Regardless of TP53 mutation status, pts with primary AML had significantly longer median OS than pts with secondary AML (no mutation: 22.0 mo [17.6-26.1] vs 6.2 mo [2.5-12.6], P < 0.001; with mutation: 8.8 mo [3.9-14.2] vs 3.8 mo [1.4-6.8], P = 0.014). VEN-based therapies were adopted more quickly by pts aged ≥ 75 y vs < 75 y (P < 0.001) and those treated at community or government centers vs academic centers (P = 0.045); no association between geographic region and time to adoption of VEN-based therapies was observed. Conclusions Overall, outcomes of pts with AML in the CONNECT® Myeloid Disease Registry are consistent with previously reported outcomes. Median OS is longer in primary AML vs secondary AML regardless of whether pts receive alloHSCT or have mutations in TP53. This analysis suggests that contrary to commonly held perceptions, site of care does not affect overall pt outcome and that VEN adoption occurs sooner at community centers. The clinically meaningful increase in OS in pts receiving alloHSCT, when compared with pts without alloHSCT, regardless of primary or secondary AML further emphasizes the importance of providing pts with the opportunity for transplant. These observations merit further study into potential differences in treatment patterns and outcomes. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Iadademstat Combination with Azacitidine Is a Safe and Effective Treatment in First Line Acute Myeloid Leukemia. Final Results of the Alice Trial

Background: Acute myeloid leukemia (AML) is a heterogenous disease with a broad spectrum of cytogenetic and molecular aberrations. Emerging targeted therapies have improved overall survival (OS) for some specific subgroups. Outcomes, however, remain poor, particularly for patients (pts) with high-risk characteristics and not eligible for intensive treatments. Combinations with hypomethylating agents are the standard of care (SoC) for unfit patients. Azacitidine (aza) plus venetoclax has shown an OS of 14.7 months (mo) and a complete remission or complete remission with incomplete hematological recovery (CR/CRi) rate of 66% (VIALE-A trial). The presence of mutations in TP53 or FLT3-ITD predicts resistance to this combination. Overall, 50% of AML pts relapse after first line treatments, generally attributed to the persistence of leukemic stem cells (LSC) and/or their clonal evolution, emphasizing the need to develop new strategies targeting LSC, to improve outcomes. One such investigational strategy is iadademstat (iada), an oral, potent and selective inhibitor of the Lysine-Specific Demethylase 1 (LSD1/KDMA1) enzyme. This enzyme works as a master regulator of transcription by a double mechanism of action: 1) removing methyl groups from histone 3 lysines (epigenetic eraser); and 2) disrupting multiprotein transcriptional complexes on genes governing cell differentiation and stemness. In myeloid cells, LSD1 provides a scaffold for assembly of the GFI1/CoREST transcriptional repressor complex, which regulates hematopoietic differentiation. Iada-induced gene expression shifts from a proliferation to differentiation signature has been preclinically characterized in AML cells (Maes, et al., Cancer Cell 2018) and in the clinic in a Ph1 study in Relapsed/Refractory AML (Salamero, et al., JCO 2020). To date, iada has been administered to more than 100 oncology pts in Ph1 and Ph2 trials, showing manageable toxicity and encouraging preliminary activity. Herein, we present the results of the Phase 2 ALICE study (EudraCT 2018-000482-36) testing the combination of iada plus aza for the frontline treatment of unfit AML patients, with a median follow-up of 6.9 mo (range 0.7-41.9) (cut-off June 30, 2022). The final data (cut-off September 30, 2022) will be presented at the ASH meeting. Methods: Unfit adult patients with AML per WHO 2017 classification who had not received previous treatment were enrolled. Two cohort doses of iada at 60 or 90μg/m2/d (PO, 5d ON, 2d OFF weekly) were explored in combination with aza (75mg/m2 SC for 7d, either d1-5, 8-9, or d1-7 every cycle) in a 28d cycle. Primary endpoints are safety, tolerability, and establishment of the recommended phase 2 dose (RP2D); secondary endpoints include remission rates, time to response (TTR), duration of responses (DoR) and OS. Additional assessments include measurable residual disease (MRD), PK/PD determinations and molecular correlations with response. Results: Baseline characteristics for 34 per protocol pts out of 36 accrued are shown in Table 1. Safety and efficacy results are detailed in Table 2. Main safety events remain consistent with previous updates, with the most frequently reported adverse reactions (AR) being platelet reduction in 53% of pts (58% of pts entered the study with grade ≥3 thrombocytopenia). Only 2 drug-related SAEs have occurred to date. Of 27 pts evaluable for efficacy, 81% responded (64% of those with CR/CRi). Responses were rapid (91% by end of C2). 71% of CR/CRi pts became transfusion independent and 82% of CR/CRi pts evaluable for MRD were negative by flow cytometry. Mean/median DoR (CR/CRi) are 9.3 mo (SE: 1.0 mo) / NE (95% CI 10.1mo, NE). Mean/median OS are 8.2 mo (SE: 0.80 mo)/9.3 mo (95% CI 11.0 mo, NE). Median OS for pts achieving CR/CRi is ~14.3 mo. Based on PK/PD, safety and efficacy, iada at 90 μg/m2/d is the RP2D for the combination. Response assessment per specific mutations shows that 3 out of 3 (3/3) with FLT3 mutations, 7/7 with RAS pathway mutations, and 6/8 pts with TP53 mutations responded. Gene expression and cell differentiation analyses are ongoing and correlations with clinical responses will be presented. Conclusion: The combination of iada with aza produces robust, rapid and durable responses in previously untreated unfit AML patients, including those with high-risk features, with a manageable toxicity profile. Further research with iada in combination with SoC treatments for AML is warranted. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Blastic Plasmacytoid Dendritic Cell Neoplasm (BPDCN): Long-Term Outcomes and Novel Observations from a Large BPDCN Cohort

Background: Blastic plasmacytoid dendritic cell neoplasm (BPDCN), is an ultra-rare, clinically aggressive, CD123-overexpressing hematologic malignancy now reclassified in the 5th edition WHO under dendritic cell and histiocytic neoplasms (Khoury J Leukemia 2022). BPDCN is characterized by involvement of 4 major compartments: skin, bone marrow, lymph nodes, and central nervous system (CNS). BPDCN has been historically associated with a lack of treatment options, dismal outcomes (median OS 8-16 months). Now, in the modern treatment era, with CD123-based and combination therapies, emerging data and a number of significant clinical/translational breakthroughs have led to progress in this rare disease field (Pemmaraju N JCO 2022; Pemmaraju N Blood Adv 2022). In this analysis, we aim to describe long-term patient characteristics and outcomes in one of the largest ongoing, single-institution, prospectively followed group of patients (pts) with BPDCN in the world. Patient Characteristics: All pts ages ≥ 18 years with a confirmed pathological diagnosis of BPDCN and treated at our institution were included in this analysis. During Oct 1998-July 2022, 118 pts with BPDCN were identified. Median age was 66 years [range 19-86 years]. 81% were male. At presentation to our institution: n=78 untreated; n=40 had prior therapy (median # prior therapies = 1 [1-5]). Bone marrow (BM) was involved in 75 (64%), skin-only in 41 (35%), lymph node-only in 2 (2%). In addition to CD4+ and CD56+ tumor immunophenotype demonstrated: TCL-1+ (77/77 tested) and CD123+ in 100%. Cytogenetics among 106 pts available: diploid (74%), complex (22%), del (12p) in 2%, miscellaneous (2%). Median BM blast count was 8% [0-95]. Historically, as there has been no standard of care therapeutic approach, a variety of frontline chemotherapy regimens were administered over time (including in this cohort: CD123-based; ALL-based; AML-based; CHOP/lymphoma-based; HMA+VEN based; other regimens). Results, overall cohort (n=118): The median follow-up time was 23 months [4.1 - 91 mo]. Median # of chemotherapy regimens was 1 [1-6]. Median overall survival (OS) was 23 mo [0.9 -158.0 mo]. 78 (66%) pts died, with the most common cause of death being multi-organ failure. Skin-only pts with BPDCN: Remarkably, in this long-term f/u, there was no statistically significant difference between skin (n=41) vs BM involved BPDCN (n=75) in terms of either OS (20 mo vs 23 mo) or CR1 (21 vs 38 mo). SCT: 40 pts (33%) received stem cell transplant (SCT) during CR1. The median OS for pts receiving SCT was statistically significantly superior at 66 mo [5.3-90.8 mo] vs a median OS for non-SCT group in CR1 (n=31) of 23 mo [4.1-158.0], p =0.003. CNS involvement BPDCN: Notably, in this longer f/u of our series, we found that a high rate, 33%, of our pts had CSF+ at any time during the course of their BPDCN, with the vast majority of cases detected on routine/asymptomatic lumbar punctures. Prior or concomitant hematologic malignancies (PCHM): The occurrence of PCHM remains quite common in BPDCN, making up 24% of our total BPDCN population, with MDS/CMML representing the most common PCHM. Next-generation Sequencing (NGS): A molecular gene panel via NGS on BM specimens was done in n=80; 60 (75%) had TET2 abnormality (standard mutation=36, standard+variant=2, variants=22), which was the most common molecular abnormality among our BPDCN cohort. There was no statistically significant difference in terms of OS or CR1 in pts with known TET2 mutations/variants vs all others/not done. The second most common mutation was ASXL1 in 38%, followed by RAS mutations in 19%. Conclusions: This longer-term analysis (med f/u ~ 2 years) of a large, ongoing cohort of pts with BPDCN reveals several important findings including continued overall similar long-term outcomes in skin-only vs BM-involved BPDCN; significant OS improvement in CR1 for younger/fit pts who are able to proceed to SCT when available; a strikingly high incidence of CSF+ in BPDCN which is a higher incidence than previously recognized, and has led to our practice-changing approach of standard incorporation of routine LP with IT chemo for all pts with BPDCN; and a notably common occurrence of PCHM in BPDCN to be aware of in both diagnostic and treatment paradigms. With median OS still only ~2 years, BPDCN remains a highly aggressive hematologic malignancy still in need of active clinical/translational research and novel therapeutic approaches.

Abstract 4118: Brain metastases (BM) and next-generation sequencing (NGS) timing: Real-world (RW) outcomes for patients (pts) with advanced non-small cell lung cancer (aNSCLC) and MET exon 14 skipping mutations (METex14)

Abstract Background: Treatment (Tx) options for METex14 NSCLC have been limited to chemotherapy (Ctx), immunotherapy (IO), and multikinase inhibitors (MKi) until FDA approvals of selective MET inhibitors (METi) from 2020. BM are a further challenge to treat. We previously reported NGS timing in relation to start of first-line Tx (1L) for pts with METex14 NSCLC in the RW. Here we provide overall survival (OS) data, including subgroup analyses for pts with known BM prior to 1L initiation. Methods: Data originating from ~280 US cancer clinics (~800 sites of care) from the US electronic health record-derived de-identified Flatiron Health Foundation Medicine clinico-genomic database (FH-FMI CGDB) were evaluated. Pts were ≥18 yrs, diagnosed with advanced/metastatic NSCLC from Jan 1, 2011, with NGS-confirmed METex14, ≥1 line of Tx, and ≥90 days follow-up from 1L initiation to cutoff date (Dec 2020). Pts treated with capmatinib or tepotinib were excluded. We examined Tx patterns and OS for pts with and without known BM before starting 1L, and OS for pts with an NGS report before or on/after 1L initiation (time from 1L initiation to death or censoring, analyzed using Kaplan-Meier method with left truncation adjustment to mitigate bias due to the prerequisite for NGS testing). HRs were estimated using Cox regression models with covariates adjustment to estimate risk of death. Results: The analysis included 156 METex14 NSCLC pts: median age 75 yrs, 59% females, 85.3% non-squamous (NSQ), 9% squamous cell carcinoma (SQ), and 64% had history of smoking. A total of 31 pts (20%) had BM before 1L initiation: median age 73 yrs, 61% females, 94% NSQ, 0% SQ, and 71% had history of smoking. Overall, treatment patterns were similar for pts with and without known BM before 1L initiation (Ctx with or without IO, IO monotherapy, and MKi were all used in 1L). Median OS for pts with BM was 8.11 mo (95% CI, 6.0-19.8) and 14.39 mo (11.1-22.1) for pts without BM, corresponding to a 73% increase in risk of death for pts with BM vs pts without BM by univariate analysis (HR 1.73 [95% CI, 1.06-2.82]; P=0.03). Median OS for pts with an NGS report before and on/after 1L initiation was 13.3 mo (8.8-26.2) and 6.8 mo (0.03-19.8), respectively (P=0.45); HR by multivariate analysis was 0.78 (0.45-1.35) when comparing pts with NGS report before vs on/after 1L initiation. Conclusions: Tx patterns in 1L were similar for pts with or without BM at time of 1L initiation; presence of BM prior to 1L was associated with inferior outcomes. Numerically longer OS was observed in pts with NGS report prior to 1L initiation vs pts with NGS report on/after 1L, supporting further investigation of underlying causes. These data underline the ongoing unmet clinical need of METex14 NSCLC BM and the importance of investigating upfront comprehensive genomic testing in NSCLC. Citation Format: Xiuning Le, Stephen Robb, Nydia Caro, Simona Doria, Whitney C. Rhodes, Wen-Hsing Wu, Julia Kim, Vincent Prêtre, Fen Ye. Brain metastases (BM) and next-generation sequencing (NGS) timing: Real-world (RW) outcomes for patients (pts) with advanced non-small cell lung cancer (aNSCLC) and MET exon 14 skipping mutations (METex14) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 4118.

Abstract 495: Outcomes of an expanded access program for patients with gastrointestinal stromal tumors in low- and middle-income countries

Abstract Background: Gastrointestinal Stromal Tumors (GIST) treatment was revolutionized by the introduction of the tyrosine kinase inhibitor imatinib. More recently, sunitinib, has prolonged survival in patients (pts) with metastatic disease resistant to imatinib. 10-year survival for high-risk pts treated with adjuvant imatinib is 79%. Median overall survival for pts with unresectable or metastatic GIST is 4.8 years. In order to bring access to these life-saving medications to pts in low- and middle- income countries (LMICs), The Max Foundation (TMF) developed its Max Access Solutions (MAS) program in 2017, after administering Novartis’ imatinib (Glivec) International Patient Assistance Program (GIPAP) since 2002. We analyzed the outcomes of pts enrolled in these programs from 2002-2020 for both adjuvant (n=2100) or metastatic (n=9,867) GIST across 67 countries. Methods: Demographic data and treatment indications were reported by treating physicians. The clinical status of each enrolled pt was reviewed with TMF every 3-4 months. Pts who were lost to follow-up (LTFU) (n=2282) with metastatic or unresectable disease were presumed to be deceased. Alternatively, for patients treated in the adjuvant setting an imputation based informed censoring model was used to estimate events for LTFU patients (n=477). Kaplan-Meier analysis was used to estimate the distribution of progression-free (PFS) and overall survival (OS). Results: The median age at diagnosis was 54 and 55 years in the adjuvant and metastatic groups, respectively. Males comprised 59.5% of all metastatic/unresectable cases and females 51.3% of adjuvant cases. Median OS for pts treated with imatinib for unresectable or metastatic disease was 5.8 years (CI: (5.6, 6.1)) and PFS was 3.5 years (CI: (3.5, 3.7)). Pts treated with imatinib in the adjuvant setting had a 10-year OS of 72% and PFS of 70%. Median OS of pts with metastatic disease treated with sunitinib was 2 years (CI: (1.5, 2.5)); PFS was 1.2 years (CI: (0.9, 1.6)). Multivariate analysis showed that male sex was a negative prognostic factor in both the metastatic and adjuvant setting. Other variables including World Bank Income groups and frequency of contact were not associated with increased mortality or progression in any group. Discussion: This is the largest known cohort of GIST outcomes and the first study to present outcomes from predominantly low- and middle-income countries. Additionally, studies in LMICs are challenged by high rates of lost to follow-up which can falsely prolong or shorten survival times due to censoring. Therefore, we employed an informed censoring model to better reflect survival in the adjuvant setting. We demonstrate that access to a targeted oral anti-cancer therapy results in outcomes similar to those in high resource countries and provides a model to help close the global gap in cancer outcomes. Citation Format: Edward Lloyd Briercheck, J. Michael Wrigglesworth, Philip Stevenson, Alicia A. Annamalay, Pat Garcia-Gonzalez, Michael Wagner. Outcomes of an expanded access program for patients with gastrointestinal stromal tumors in low- and middle-income countries [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 495.

Personalized circulating tumor DNA (ctDNA) analysis in patients with recurrent/metastatic head and neck squamous cell cancer (R/M HNSCC).

6052 Background: Immuno-oncology agents (IO) have become standard-of-care in the treatment of R/M HNSCC, but only a subset of patients (pts) benefit. Highly sensitive quantification of plasma circulating tumor DNA (ctDNA) may permit real time assessment of disease under selective pressures of treatment. Methods: R/M HNSCC pts treated with platinum-based chemotherapy (CT) or IO (anti-PD1/L1 +/- second IO) underwent serial ctDNA collection pre-cycles 1/2/3 and at disease progression, corresponding to timepoints (T) 1-4. T1 was considered baseline. Whole exome sequencing of pt tumor tissue identified patient specific somatic variants which were used as targets for RaDaR, a personalized multiplexed PCR-based NGS assay. Matched buffy coat DNA was sequenced to filter germline mutations and identify confounding CHIP. RaDaR was applied at each available T, an estimated variant allele frequency (eVAF) was calculated and correlated with progression free- (PFS) and overall- survival (OS). Findings were compared against prior (ESMO 2021) data generated using a fixed 580 gene CAPP-seq (CAncer Personalized Profiling by deep Sequencing) panel designed specifically against squamous cell carcinoma. Results: A total 114 plasma samples from 38 pts were analyzed. Of 35 pts with ctDNA detected at T1 and/or T2, 26 received IO and 9 CT. Median age was 62 (20-84), 77% were male, 69% prior smokers and 26% HPV positive. Median PFS and OS, for all 35 pts, was 2.57 mo (95% CI 0.48-4.66) and 8.37 mo (95% CI 5.42-11.32) respectively. For IO treated pts, median PFS was 2.45 mo (95% CI 0-5.18) and median OS 7.38 mo (95% CI 3.84-10.93). RaDaR panels targeted a median 48 variants (17-50). ctDNA was detected in 35/38 (92%) patients at baseline, with median eVAF 0.345% (range 0.0004% - 43.37%). ctDNA abundance at baseline did not correlate with PFS or OS. A decrease in Δ eVAF from T1 to T2, by &gt; 30%, or &gt; 50% identified pts with improved PFS, with HR 0.45 (0.21, 0.96) p = 0.04, 0.31 (0.14, 0.70) p &lt; 0.01, and 0.23 (0.10, 0.56) p &lt; 0.01, respectively. Similar results were observed for the 26 IO pts, with HR 0.40 (0.16, 1.03) p = 0.06, 0.19 (0.05, 0.66) p &lt; 0.01, and 0.06 (0.01, 0.47) p &lt; 0.01, respectively. A similar, but non-significant, trend was seen in median OS for pts with a decrease vs. increase in Δ eVAF (T1 to 2), 8.8 mo vs 7.3 mo (HR = 0.87 (0.42, 1.79)). For 31 pts, a comparison of Δ ctDNA levels, based on personalized RaDaR vs. CAPP-seq assays, from T1 to T2 demonstrated a correlation coefficient of R = 0.57, P &lt; 0.01. Conclusions: In pts with R/M HNSCC, a decrease in ctDNA eVAF after first treatment correlated with improved PFS. There was a significant correlation between fixed CAPP-seq and personalized RaDaR assays when comparing Δ in ctDNA levels. Clinical Trial: NCT03712566.

Clinical and molecular features of low prostate-specific membrane antigen (PSMA) expression in patients (pts) with metastatic castration resistant prostate cancer (mCRPC).

167 Background: Low PSMA uptake on positron-emission tomography is seen in up to 30% of mCRPC pts and represents a clinically distinct subgroup with adverse outcomes. We assessed transcriptional and clinical features associated with low PSMA ( FOLH1) gene expression in mCRPC. Methods: A retrospective analysis of mCRPC biopsy samples with RNA-seq data was undertaken. Normalized FOLH1 expression was compared across histologic subtypes and sites of disease. We assessed the association between FOLH1 expression, selected androgen receptor (AR) target genes, master regulators of neuroendocrine differentiation, and previously validated AR activity and treatment-associated small cell neuroendocrine carcinoma (t-SCNC) transcriptional signature scores using Pearson correlations. Associations between FOLH1 and both PSA50 response to subsequent AR-targeted therapy and overall survival (OS) were examined by logistic regression and Cox proportional hazard models, respectively. Results: Samples from 97 pts were identified, of which 18% harbored t-SCNC histology. 45% of pts had visceral metastases at the time of biopsy, and 41% received subsequent AR-targeted therapy. Median FOLH1 expression was lower in pts with visceral metastases vs no visceral metastases (14.7 vs 15.6, p = 0.02) but was not significantly different across t-SCNC vs adenocarcinoma biopsies (14.3 vs 15.4, p = 0.13). FOLH1 expression was positively correlated with AR transcriptional activity and AR target genes, and negatively correlated with master regulators of neuroendocrine differentiation and t-SCNC transcriptional signature scores (Table). Low FOLH1 expression did not predict PSA50 response to subsequent AR-targeted therapy (OR 0.97, p = 0.8), but was associated with shorter OS on univariate analysis (HR 1.09, 95% CI 1.02-1.16, p=0.01). A post-hoc analysis revealed a trend towards decreased median OS in pts with FOLH1 expression &lt;12 (7.5 vs 17.1 months, log-rank p = 0.06). Conclusions: In this retrospective analysis of mCRPC pts, low FOLH1 expression was associated with transcriptional features of t-SCNC, decreased AR activity, and shorter OS. These findings are hypothesis-generating and prospective validation is needed.[Table: see text]

Abstract P5-16-15: Post-progression therapy outcomes in patients (pts) from the phase 3 ASCENT study of sacituzumab govitecan (SG) in metastatic triple-negative breast cancer (mTNBC)

Abstract Background: SG is an antibody-drug conjugate composed of an anti-Trop-2 antibody coupled to the cytotoxic SN-38 payload via a proprietary, hydrolyzable linker. SG is FDA approved for pts with mTNBC who received ≥2 prior chemotherapies (≥1 in the metastatic setting). The confirmatory phase 3 ASCENT study (NCT02574455) in pts treated in second line or greater (2L+) mTNBC demonstrated significant progression-free survival (PFS) and overall survival (OS) benefit of SG over single-agent chemotherapy treatment of physician’s choice (median PFS: 4.8 vs 1.7 months, HR 0.43, P&amp;lt;0.001; median OS: 11.8 vs 6.9 months, HR 0.51) in the full trial population, with a manageable safety profile. However, outcomes and patterns of subsequent therapy for pts who discontinue SG following progressive disease (PD) are not well characterized. This post hoc subgroup analysis investigates post-progression treatment and OS of pts who discontinued SG due to PD during the ASCENT trial. Methods: In ASCENT, pts with mTNBC refractory/relapsing after ≥2 prior chemotherapies (≥1 in the metastatic setting) were randomized 1:1 to receive SG (10 mg/kg IV on d 1 and 8, every 21 d) or TPC (capecitabine, eribulin, vinorelbine, or gemcitabine) until unacceptable toxicity or progression. Post-progression outcomes were assessed in pts who discontinued SG due to progressive disease (PD). Pts were followed every 4 weeks for OS, including documentation of further therapy for breast cancer. Time to post-progression therapy was defined as the number of months from time of randomization until the initiation of subsequent anticancer therapy. OS was analyzed in pts who received post-progression therapy vs those who did not and defined as the number of months from randomization or from end of SG treatment, using Kaplan Meier estimates and Cox regression. Results: 222/267 (83%) pts who were randomized to receive SG discontinued SG due to PD. In these patients median age was 53 years (range, 27-82), median number of prior anticancer regimens was 4, and 7% had known germline BRCA1/2 mutations. Pts received SG for a median duration of 4.2 months (range, 0.0-18.7). Following SG discontinuation, post-progression therapy was received by 73% (n=163) of pts; common post-SG therapies included eribulin (n=70; 32%), carboplatin (n=34; 15%), capecitabine (n=34; 15%), and atezolizumab, (n=15; 7%). The median time to receipt of post-progression therapy was 5.4 months (range, 1.0-19.8). Median OS in pts who received any post PD treatment vs those who did not receive post PD treatment following SG was 13.4 vs 7.3 months (HR, 0.46; 95% CI, 0.32-0.67; P&amp;lt;0.0001) from time of randomization and 7.9 vs 2.0 months (HR, 0.14; 95% CI, 0.09-0.22; P&amp;lt;0.0001) from end of SG treatment, respectively. In pts who received eribulin, carboplatin, atezolizumab, or capecitabine, median OS was 14.1 (95% CI, 10.9-14.9), 13.6 (95% CI, 10.6-15.9), 16.5 (95% CI, 8.7 to not evaluable), and 14.9 (95% CI, 10.9-16.8) months from time of randomization and 8.4 (95% CI, 6.8-9.2), 8.9 (95% CI, 6.7-10.8), 8.6 (95% CI, 4.3 to not evaluable), and 8.9 (95% CI, 6.6-10.3) months from end of SG treatment, respectively. Conclusions: In ASCENT, the majority of pts who discontinued SG due to PD were able to receive subsequent therapy post-progression. Pts who received post PD therapy following SG had significantly improved median OS over those who did not receive further therapy. Pts who received eribulin, carboplatin, atezolizumab, or capecitabine, as post PD therapy had similar median OS. These results indicate that treatment with SG does not prevent receipt of further systemic therapy. Citation Format: Javier Cortés, Aditya Bardia, Delphine Loirat, Sara M. Tolaney, Kevin Punie, Mafalda Oliveira, Sara A. Hurvitz, Adam Brufsky, Sagar Sardesai, Kevin M Kalinsky, Tiffany Traina, Erika Hamilton, Joyce O’Shaughnessy, Véronique Diéras, Lisa A. Carey, Martine Piccart, Sibylle Loibl, Hope S. Rugo, Yanni Zhu, See Phan, Luca Gianni. Post-progression therapy outcomes in patients (pts) from the phase 3 ASCENT study of sacituzumab govitecan (SG) in metastatic triple-negative breast cancer (mTNBC) [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P5-16-15.

Efficacy analysis of regorafenib combined with PD-1 inhibitors in elderly patients with metastatic colorectal cancer.

105 Background: Recently, an increasing number of patients (pts) with advanced colorectal cancer tend to receive regorafenib plus PD-1 inhibitors. However, there is a lack of information regarding real world effects of this therapeutic strategy, especially in elderly Chinese pts. We aimed to investigate the treatment patterns, clinical outcomes and prognostic factors of regorafenib plus PD-1 inhibitors therapy to Chinese elderly pts with advanced colorectal cancer. Methods: Retrospective analysis of a cohort of pts with advanced colorectal cancer received treatment of regorafenib combined with PD-1 inhibitors enrolled in our institution between January 2019 and January 2021. Among them, pts with 60 years or older were included in our final analysis. Overall survival (OS) and progression free survival (PFS) were estimated by Kaplan Meier curves, and other endpoints included objective response rate, prognostic factors. Results: At a median follow-up of 16.2 months, a total of 24 pts were enrolled in a real-world dataset from Henan Cancer Hospital. The median age was 68 years; 62.5% were female. The median OS and PFS were 15.03 months (95% CI 7.0-23.0) and 4.0 months (95% CI 1.8-6.2), respectively. 2/24 partial responses (PR) observed (ORR = 8.3%); Stable disease (SD) was obtained in 15/24. Disease control rate (PR plus SD) was therefore obtained in 17/24 (70.8%). When compared with PFS and OS in different initial daily doses group of regorafenib (such as ≤80, 120 or 160 mg), we concluded that there were no significantly difference between these groups. And similar results were found in final daily doses groups. However, it showed a trend toward better PFS in pts with ≥120 mg final daily doses group compared to &lt;120 mg group (median PFS was 3.5 months in &lt;120mg group versus 10.0 months in ≥120mg group). Median OS of pts for male gender demonstrated increasing trends compared with those female gender (20.2 months in males versus 13.2 months in females). Pts with previous treated with regorafenib had longer median OS than those without it (20.2 months versus 13.2 months). There were no obvious correlation for survival between pts only had liver metastasis and lung metastasis (P&gt;0.05). No significant associations for survival could be seen in other tumor sites, K-RAS status, any kind of PD-1 inhibitors and whether previously treated with bevacizumab (P&gt;0.05). Conclusions: This real-world dataset confirms that Chinese elderly pts with advanced colorectal cancer also can benefit from treatment of regorafenib combined with PD-1 inhibitors, similarly with this combination therapy strategies in all age pts. Furthermore, a longer survival time were obtained in pts with prior treated with regorafenib. The result also suggests that pts with ≥120mg final daily doses group had better PFS. However, further larger cohorts research should investigate whether the PFS advantage in high dose group could eventually lead to improved OS outcomes.