Abstract P4-01-04: ESR1 mutation in cell free DNA (cfDNA) is associated with significantly increased circulating tumor cell (CTC)-clusters and progress in stage III/IV breast cancer after systemic treatments

Introduction: CTCs play a critical role in the process of tumor metastasis, and a portion of CTCs may form clusters that contain two or more CTCs bound together which were reported to have up to 50-fold of potential of forming distant metastasis in breast cancer (MBC) as compared to individual CTCs. However, molecular and genomic characterization of CTCs cluster remain largely unknown. Here we report a highly significant correlation between ESR1 mutation in cfDNA, CTCs count and CTC-cluster, which may help to understand MBC metastasis and predict treatment benefit, especially for metastatic or recurrent disease.

Methods: A total of 80 whole blood samples (7.5ml/each) were collected from 80 patients with stage III/IV BCa after informed consent under IRB-approved trial at the RHLCCC at Northwestern University before and after systemic therapies. Among these 80 patients, 41 patients received chemotherapy and 23 patients received endocrine therapy, among which 20 patients received combo treatments (16 plus Palpociclib, 1 plus Ribociclib, 2 plus Everolimus, and 1 plus trastuzumab). CTC enrichment and enumeration were performed in CELLTRACKS ANALYZERII® System (Menarini) by using CTC Kit Meanwhile, we detected the ESR1 hotspot mutations (Y537S and D538G) in plasma cfDNA from all 80 patients by Droplet digital PCR (ddPCR) assay using the QX200 ddPCR System (Bio-Rad). cfDNA was isolated from 2 mL of plasma using the QIAamp Circulating Nucleic Acid Kit (Qiagen) and the MAF was analyzed using QuantaSoft software (Bio-Rad).Database of CTCs and ESR1 mutation was linked with clinical database. Kruskal-Wallis test was used for statistics.

Results: Of the 80 samples analyzed, there were 57 samples without ESR1 mutations (Group 1), and 23 samples that had ESR1 mutations (8 Y537S mutations and 23 D538G mutations, Group 2). CTC positive (≥5) were detected in 13/57 samples (Group 1) and 15/23 samples, and the average amounts of CTCs were 21.77 CTCs/each sample and 59.86 CTCs/each sample in Group 1 and Group 2 respectively. There was a significant association between ESR1 mutations and high level of CTCs (P=0.000088). More important, CTC-clusters were found in 3 samples in Group 1 (5.26%) and in 5 samples in Group 2 (21.74%) respectively. There was a significant correlation between ESR1 mutations and CTC-clusters (P=0.026). Furthermore, there were 18/57 patients in group 1 and 5/23 in group 2 receiving chemotherapy. Moreover, 26/57 in group 1 and 15/23 in group 2 that received chemotherapy. Our results also confirmed that both endocrine therapy and chemotherapy benefited more patients without ESR1 mutations in compared with patients with ESR1 mutations (P<0.05).

Conclusion:We first elucidated the association between ESR1 mutations in ctDNA and CTC-cluster in MBC patients, and provides new insights on the molecular mechanisms associated with the metastasis process. In addition with the highly significant association between ctDNA ESR1 mutations and endocrine resistance we describe a new association allowing to expand the prognostic and predictive role of both tests enabling monitoring the metastatic prognosis and endocrine resistance for clinical decision-making.

Citation Format: Zhang Q, Gerratana L, Zhang Y, Flaum L, Shah A, Davis A, Behdad A, Gradishar W, Platanias L, Cristofanilli M. ESR1 mutation in cell free DNA (cfDNA) is associated with significantly increased circulating tumor cell (CTC)-clusters and progress in stage III/IV breast cancer after systemic treatments [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-01-04.

Abstract P5-17-03: How is inflammatory breast cancer (IBC) different? Integration of clinico-pathological features and circulating tumor cells (CTCs)-based biomarkers for disease and prognostic assessment

Background: Since IBC is rare and burdened by a particularly unfavorable prognosis, biomarkers able to enhance diagnosis and risk assessment are of pivotal importance and a current unmet need. The aim of this study is to integrate standard clinico-pathological features with CTCs-based biomarkers for a more objective and detailed characterization of IBC.

Methods: This study analyzed retrospectively 251 Advanced Breast Cancer (BC) patients (pts) longitudinally characterized for CTCs and CTCs-based biomarkers at Thomas Jefferson University (Philadephia, PA) and Northwestern University (Chicago. IL). CTCs were enumerated through the CellSearch system (Menarini Silicon Biosystems), and characterized for HER2 expression using the CellSearch CXC Kit. Pts were defined as stage IV aggressive based on the previously reported ≥5 CTCs cut-off (Davis et al. 2018). Associations between clinical features, CTC-derived biomarkers and IBC were tested through uni and multivariate logistic regression. Survival was tested though log-rank test.

Results: Within the analyzed cases, 46% were diagnosed with IBC and among them, 38% was stage IV aggressive. CTC clusters (CTC_CL) were detectable in 12.5% of pts and HER2 positive CTCs (HER2_CTC) in 29.5%. Notably, IBC patients (pts) had a significantly lower CTC count with respect to non-IBC (median 2.5 vs 0 respectively for non-IBC and IBC; P=0.019). BC subtype (HER2 positive BC: OR 2.97; Triple negative BC: OR 2.13), liver and bone involvement (liver: OR 0.46; bone involvement: OR 0.31) were the only significant clinico-pathological features associated with IBC at univariate logistic regression. Interestingly, a marginal significance was observed for soft tissue involvement (OR 1.65, 95%CI 0.95 - 2.87, P=0.07). Stage IV aggressive and presence of HER2_CTC at baseline were moreover inversely associated with IBC. The multivariate model confirmed the significant association between IBC and HER2 positive BC subtype (OR 2.64, 95%CI 1.08 - 6.48, P=0.034), absence of bone involvement (OR 0.31, 95%CI 0.14 - 0.68, P=0.003) and absence of HER2_CTC (OR 0.38, 95%CI 0.15 - 0.98, P=0.045). The baseline detection of CTC_CL was a strong predictor of prognosis for OS in IBC pts (median OS (mOS) 7.6 months (mts) vs not reached (NR) respectively for detectable vs non-detectable CTC_CL; P<0.0001), while a trend was observed for HER2_CTC (mOS 9.9 mts vs NR respectively for detectable vs non-detectable HER2_CTC; P<0.082). Pts negative for CTC_CL at baseline had higher odds of developing CTC_CL in later time-points if stage IV aggressive (OR 12.27, 95%CI 2.10 - 71.57, P=0.005). Despite no baseline factors were significantly associated with the onset of HER2_CTC in later time-points, a trend (P=0.05) was observed for patients without lymph node involvement (OR: 5) and with bone involvement (OR: 4.3).

Conclusion: HER2_CTC and in particular CTC_CL are promising prognostic predictors in IBC. Stage IV aggressive IBC pts could benefit from a longitudinal CTCs assessment, being more prone to develop CTC_CL and therefore at higher risk of rapid disease progression. Probably due to the tropism for soft tissue, IBC is characterized by a lower number of HER2_CTC.

Citation Format: Gerratana L, Zhang Q, Wang C, Shah A, Davis AA, Ye Z, Zhang Y, Abu-Khalaf M, Flaum L, Strickland K, Rossi G, Behdad A, Gradishar W, Platanias L, Yang H, Cristofanilli M. How is inflammatory breast cancer (IBC) different? Integration of clinico-pathological features and circulating tumor cells (CTCs)-based biomarkers for disease and prognostic assessment [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-17-03.

Abstract P3-01-08: HER2-negative metastatic breast cancer with HER2-positive circulating tumor cells (CTCs): A new CTC-defined HER2-positive subgroup

Introduction: CTCs can overexpress HER2 discordant from tumor HER2 expression. We aimed to describe characteristics of a CTC-defined group of pts with metastatic breast cancer (MBC) that is tumor HER2- and CTC HER2+ (HER2 tumor- CTC+).

Methods: We retrospectively analyzed data from pts treated at Northwestern University who had serial evaluation of CTCs and circulating tumor DNA (ctDNA). We included pts with pathologically confirmed HER2- MBC and HER2+ CTCs. CTCs were enumerated with the CellSearch immunomagnetic kit (Menarini Silicon Biosystems), HER2 expression on CTCs was determined using the CellSearch CXC Kit in 7.5 cc whole blood, and ctDNA was analyzed using the Guardant360 NGS assay (Guardant Health).

Results: Among 98 pts with HER2- MBC and CTC analysis, 46 (47%) had at least 1 HER2+ CTC. In this cohort the median age was 53. At initial BC diagnosis, 80% had early stage or locally advanced BC and 20% had de-novo metastatic disease. Baseline histology was 65% ductal, 20% lobular, 2% mixed ductal and lobular, and 13% unknown. Pathology of metastatic tumor was hormone receptor positive (HR+)/HER2- in 78% and triple negative in 22%. Detailed HER2 immunohistochemistry (IHC) and FISH results from metastases were available from 63% of pts of whom 72% had an IHC score of 0 or 1 and 28% had an IHC score of 2 with negative FISH testing. The median time from the most recent pathologic metastatic tumor assessment to the detection of a HER2+ CTC was 6.5 mo. Twenty-two pts had simultaneous (within 8 weeks) HER2- tumor confirmation and HER2+ CTC detection. The median lines of endocrine therapy (ET) for MBC prior to detection of HER2+ CTCs was 1 (range 0-5, 41% no ET, 17% 1 line, 41% >2 lines). Pts received a median of 2 (range 0-10) prior systemic therapies for MBC prior to detection of HER2+ CTCs, (20% 0 lines, 41% 1-3 lines, and 39% >4 lines). Among these 46 pts, CTCs were analyzed longitudinally in 104 samples, with HER2+ CTCs detected in 77 samples. Number of HER2+ CTCs at initial detection ranged from <5 in 24%, 5-50 in 43%, and >50 in 33%, with a median of 11.5 HER2+ CTCs. CTC clusters were noted in 37% of pts. At initial detection the proportion of CTCs that were HER2+ was 0-25% in 13% of pts, 26-50% in 46% of pts, and 51-100% in 41% of pts. Seven pts had ERBB2 aberrations in ctDNA. Of 12 pts with tumor sequencing, 2 had ERBB2 mutations, 1 had ERBB3 amplification, and 1 had overexpression of ERBB3 RNA. After detection of HER2+ CTCs, 18 pts received HER2 directed therapy (with chemotherapy in 13 pts, with endocrine therapy in 4 pts, and as monotherapy in 1 pt). Imaging demonstrated a partial response or stable disease in 9 pts (clinical benefit rate 50%), including in 1 pt with trastuzumab monotherapy, progressive disease in 8 pts, and not evaluated in 1 pt.

Conclusions: HER2+ CTCs are frequently detected simultaneously or soon after HER2- tumor assessment in MBC. Within this newly defined subgroup, the several responses seen with HER2 targeted therapy serve as a proof of concept that HER2 tumor- CTC+ patients can benefit from HER2 targeted therapy. Future studies are needed to determine a clinically relevant threshold for HER2+ CTCs to guide further study of HER2 therapy combinations in HER2 tumor- CTC+ pts.

Citation Format: Shah AN, Gerratana L, Zhang Q, Davis AA, Zhang Y, Flaum L, Behdad A, Platanias L, Gradishar WJ, Cristofanilli M. HER2-negative metastatic breast cancer with HER2-positive circulating tumor cells (CTCs): A new CTC-defined HER2-positive subgroup [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-01-08.

Abstract P4-01-18: Correlation between circulating tumor DNA (ctDNA) alterations and circulating tumor cells (CTC) uncovers new mechanisms of metastasis for patients with metastatic breast carcinoma (MBC)

Introduction:Novel molecular diagnostics including CTCs and ctDNA have been proved to predict disease metastasis and survival. However, the frequency of detection of actionable mutations using CTCs and ctDNA is variable based upon tumor related factors and diagnostic platform sensitivity. Herein, we evaluated a novel NGS technology in the ability of detecting driver and clonal genomic abnormalities in samples from MBC patients, and compared ctDNA alterations with CTCs and CTC-cluster. This study demonstrated several novel correlation between some specific ctDNA alterations and CTCs or CTCs related biomarkers, which opened new insight on mechanisms of metastasis for MBC.

Methods: This study included 52 samples from 26 patients with stage III/IV BCa treated at NMH (2016-2017) and who received standard systemic treatments based on disease subtypes. Whole blood samples (7.5ml/each) were used for CTC enrichment and enumeration in FDA approved CELLTRACKS ANALYZERII® System (Menarini). ctDNA from clinical plasma samples was analyzed by using PredicinePLUS, a NGS-based assay (Predicine Inc) with a 180-gene panel for genomic alterations mutations. Results of CTCs and ctDNA alterations were linked to clinical database. Matched pairs variations between CTCs and ctDNA alterations was compared by Wilcoxon signed-ranks test and Kruskal-Wallis test.

Results: Genomic Alterations (SNVs, Indels and copy number variations) were detected on 52 genes by PredicinePLUS assay. All samples (100%) demonstrated at least 1 somatic alterations. There were 75 mutations detected within 29 genes, and the variant frequency of mutated genes ranges from 0.11% to 68.56%. Increased CTCs were highly significantly correlated with genomic alterations in the genes (wild type vs alterations) including GATA3 (8vs 37), ESR1 ( 2.5 vs 41.3), CDH1 (3.5 vs 50.5) and CCND1 (4 vs 120) (P<0.01). Decreased CTCs were correlated with alterations of CDKN2A (20.5 vs 0) (P=0.025). CTC-cluster appear associated predominantly with alterations of CDH1 (P=0.0018), CCND1 (P=0.008) and BRCA1 (P=0.04). Furthermore, in HER positive CTCs group, ERBB2 mutations caused increased CTCs in compared with ERBB2 wild type (0 vs 5), when CCND1, CDKN2A, GATA3 and TP53 alterations were associated with increase of HER2 negative CTCs.

Conclusions: By using the novel diagnostic platform with the ability to identify ctDNA mutation and copy number variation, this study demonstrated several novel genes alterations which were highly correlated with CTCs, CTC-cluster and HER2. Some genes (CCND1 and CDH1) got involved into the changes on both CTCs and CTC-cluster, when some genes (CCND1, CDKN2A, ESR1 and GATA3) were related with change of CTCs and HER2 expression. Correlation of CTCs and ctDNA can be reliably and routinely used as non-invasive method for monitoring disease metastasis and predict the prognosis in MBC in clinic.

Citation Format: Davis A, Zhang Q, Gerratana L, Zhang Y, Flaum L, Shad A, Behdad A, Gradishar W, Platanias L, Cristofanilli M. Correlation between circulating tumor DNA (ctDNA) alterations and circulating tumor cells (CTC) uncovers new mechanisms of metastasis for patients with metastatic breast carcinoma (MBC) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-01-18.

Abstract P4-01-08: Characterization of circulating tumor free DNA (ctDNA) obtained from patients with metastatic breast carcinoma (MBC) undergoing systemic therapies using comprehensive genomic profiling

Introduction:Therapeutic benefit from individual treatments in patients with MBC is limited to small subsets of patients and of short duration due to tumor heterogeneity. Novel molecular diagnostics including ctDNA has been shown to predict response or resistance and survival. However, the frequency of detection of actionable mutations using ctDNA is variable based upon tumor related factors and diagnostic platform sensitivity (e.g. ddPCR or NGS). We evaluated a novel NGS technology in the ability of detecting driver and clonal genomic abnormalities in samples from MBC patients. Moreover, we wanted to compare the new technology to another state-of-the-art, commercially available diagnostic ctDNA testing with similar sensitivity to demonstrate both are able to detect genomic abnormalities in MBC.

Methods: This study included 30 samples from 15 patients with stage III/IV BCa treated at NMH (2016-2017) and who received standard systemic treatments based on disease subtypes longitudinally characterized for ctDNA before or 3 months after systemic therapies respectively. ctDNA from clinical plasma samples was first analyzed using PredicinePLUS, a NGS-based assay (Predicine Inc) with a 180-gene panel for genomic alterations mutations. The results were then independently analyzed with Guardant360TM (Guardant Health), a 73-gene panel. Matched pairs variations between Guardant360TM and Predicine was compared by Wilcoxon signed-ranks test. The prognostic impact of ctDNA was tested through Cox regression.

Results: Genomic Alterations (SNVs, Indels and copy number variations) were detected on 43 genes by PredicinePLUS assay. All samples (100%) demonstrated at least 1 somatic alterations. There were 75 mutations detected within 29 genes, and the variant frequency of mutated genes ranges from 0.11% to 68.56%. Median variant frequency was around 3.42%. Key cancer related genes including TP53, ESR1, PIK3CA, PTEN and BRCA1, are frequently mutated. Copy number variation were detected on 18 genes, among which 15 genes showed copy number gain, including MYC, PIK3CA, CCND1, and 3 genes (ATM, BRCA1 and CDKN2A) with copy number loss. There were no significant difference of % ctDNA (P=0.3967) and number of variations (P=0.5) between results of Predicine and Guardant360TM, neither to the comparison of main detected alterations (BRCA1, ESR1, MYC, PIK3CA and TP53) with Guardant360TM and Predicine (P=1). Furthermore, results from Predicine indicated that there is correlation with treatment response and benefit. A significant decrease on variations in %ctDNA levels (P=0.028) and variations in the number of genomic variants (P=0.028) after systemic therapies, was associated with longer overall survival.

Conclusions: Our study describes a novel diagnostic platform with the ability to identify ctDNA mutation and copy number variations in patients with MBC receiving systemic therapy. We also confirm that when comparing ctDNA using NGS platforms with similar sensitivity, the results are robust and reproducible which indicates that these technologies can be reliably and routinely used as non-invasive method for monitoring response to systemic therapies and predict the prognosis in MBC.

Citation Format: Zhang Q, Gerratana L, Zhang Y, Flaum L, Gradishar W, Platanias L, Cristofanilli M. Characterization of circulating tumor free DNA (ctDNA) obtained from patients with metastatic breast carcinoma (MBC) undergoing systemic therapies using comprehensive genomic profiling [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-01-08.

Abstract P3-01-19: HER2-positive circulating tumor cells (CTCs) in advanced breast cancer (BC): A feature independent of BC subtype

Introduction: HER2 overexpression is observed on CTCs in advanced BC (ABC), but their significance is not known. We aimed to describe clinical, pathologic, and molecular associations with HER2 overexpression on CTCs in ABC patients (pts).

Methods: We conducted a retrospective analysis of data from ABC pts treated at Thomas Jefferson University and Northwestern University who had evaluation of CTCs and circulating tumor DNA (ctDNA). CTCs were enumerated with the CellSearch immunomagnetic kit (Menarini Silicon Biosystems), HER2 expression on CTCs was evaluated using the CellSearch CXC Kit, and ctDNA was analyzed using the Guardant360 NGS assay (Guardant Health). Associations with the presence of HER2+ CTCs were explored through univariate and multivariate logistic regression. Kruskal-Wallis testing evaluating HER2+ CTCs as a continuous variable was also conducted to confirm consistency of findings. Time to development of HER2+ CTCs was evaluated using Cox proportional hazards regression analysis.

Results: Baseline CTCs were evaluated in 209 pts (10% stage III, 90% stage IV) of whom 41% had no detectable CTCs, 23% had 1-4 CTCs, and 36% had >5 CTCs (stage IV aggressive). Twelve percent had CTC clusters. At least 1 HER2+ CTC was seen in 33% of pts at baseline draw. Of 39 patients with HER2+ BC, only 18% had HER2+ CTCs. Of patients with HER2+ CTCs, 55% had hormone receptor positive BC, 28% had triple negative BC, and 18% had HER2+ BC. On univariate logistic analysis, BC subtype or HER2 status was not associated with the presence of HER2+ CTCs. IBC pts represented 52% of pts and were less likely to have HER2+ CTCs (OR 0.40 95% CI 0.19-0.84). Bone metastases were associated with an increased likelihood of HER2+ CTCs (OR 2.46, 95% CI 1.12-5.38); however, other sites of metastases and number of metastatic sites were not correlated with HER2+ CTCs. Aggressive disease features including >5 CTCs and presence of CTC clusters were strongly associated with HER2+ CTCs (OR 15.72, 95% CI 6.89-35.8 and 8.97, 95% CI 3.23-24.89, respectively). Of 168 pts with ctDNA analysis, ERRB2 aberrations were seen in 22% of pts and were significantly associated with HER2+ CTCs (OR of 3.74, 95% CI 1.45-9.63). On multivariate analysis, the associations with >5 CTCs and ERBB2 alterations in ctDNA remained statistically significant. The associations of HER2+ CTCs with bone disease, >5 CTCs, CTC clusters, and ERBB2 alterations in ctDNA, and the inverse relationship with IBC were consistent when HER2+ CTCs were evaluated as a continuous variable with Kruskal-Wallis testing. Among pts without HER2+ CTCs at baseline, the time to detection of HER2+ CTCs correlated with the presence of bone metastases (HR 3.40, 95% CI 1.14-10.19), >5 CTCs (3.77, 95% CI 1.33-10.70), and visceral disease (HR 3.00, 95% CI 1.07-8.44).

Conclusions: HER2+ CTCs are common in ABC, independent of HER2 status of the tumor, and, in fact, common in the luminal BC. HER2+ CTCs were also strongly associated with CTC characteristics of aggressive disease with poor survival (CTCs clusters and >5 CTCs) and ERBB2 aberrations in ctDNA. Further studies will be investigating the role of HER2+ CTCs in endocrine resistance and the potential of anti-HER2 therapy in this unique CTC-defined setting.

Citation Format: Shah AN, Gerratana L, Davis AA, Zhang Q, Zhang Y, Rossi G, Wang C, Strickland K, Yang H, Flaum L, Abu-Khalaf M, Behdad A, Ye Z, Platanias L, Gradishar WJ, Cristofanilli M. HER2-positive circulating tumor cells (CTCs) in advanced breast cancer (BC): A feature independent of BC subtype [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-01-19.

Abstract P3-01-10: Associations between plasma Interleukin 2 (IL-2) and HER2 expression in circulating tumor cell (CTC) and MYC alterations in circulation tumor DNA (ctDNA) open a new insight on immune microenvironment for patients with metastatic breast cancer (MBC)

Introduction: Overexpression of HER2 has been reported to be associated with metastasis and poor prognosis of patients with MBC. We reported in AACR 2018 that HER2 overexpression is associated with CTC-cluster. Preclinical data suggested that MYC and HER2 cooperate to drive stem cell phenotype and poor prognosis in MBC (Nair R). Furthermore, IL-2 upregulates the transcription of MYC (Grigorieva I) and gets involved into its alterations. We reasoned that further understanding of interactions of HER2 in CTC and MYC will be important to elucidate the mechanism of metastasis of MBC. Herein, we report a significant correlation between the plasma IL-2 level and HER2 expression in CTCs, and the IL-2 related MYC ctDNA alterations in MBC.

Methods: This study enrolled 43patients with stage III/IV BCa at the Northwestern Memorial Hospital (2016-2017) that had longitudinally detection of CTCs and ctDNA. Whole blood samples (7.5ml/each) were collected for CTCs enumeration by using CELLTRACKS ANALYZERII® System (Menarini) contains antibodies of anti-EpCAM for capturing CTCs, anti-CK-PE for epithelial cells, DAPI for nucleus, anti-CD45-APC for leukocytes and anti-HER-2/neu-FLU. The CTCs were classified based on phenotype as CK+, EpCAM+, DAPI+ and CD45-.Plasma ctDNA was analyzed using the Guardant360 TM NGS-based assay (Guardant Health), a 73 genes panel. ELISA for IL-2 was performed by using patients' plasma. Database of IL-2, HER2, CTCs and ctDNA was linked with clinical database and analyzed by Kruskal-Wallis test.

Results: CTCs ≥ 5 were found in 20 patients (46%). There were 15 patients that had HER2 negative CTCs (Group 1), and 5 patients had HER2 positive (Group 2) CTCs. The level of IL-2 was much higher in Group 1 (88.17pg) compared to Group 2 (66.81pg), indicating that patients with HER2 positive CTCs have significant lower IL-2 than patients with negative CTCs (P=0.02). Meanwhile, ctDNA MYC alterations were detected in 10 patients (including 1 L114R mutation, 7 CNV and 2 SNV) who have the average IL-2 level as 94.00pg. There were 11 patients without any alterations of MYC had average IL-2 level of 70.17pg, which indicated that patients with alterations in the ctDNA MYC have significant higher level of IL-2 in compared with patients without MYC alterations (P=0.02).

Conclusions: Findings of the correlation between overexpression of HER2 in CTCs and low IL-2 level indicated that low immunity may contribute to more aggressive MBC. And the higher level of IL-2 appear associated predominantly with MYC genomic alterations indicated that overexpression of MYC may also stimulate the immune response by upregulating IL-2 via a reverse feedback pathway. We postulated that increasing IL-2 suppresses the HER2 expression in CTC and breaks cooperation between HER2 and MYC. Although the interactions between them still unknown, our results suggest that IL-2 related immune microenvironment acts as a key player to suppress HER2- and MYC-mediated progress in MBC, including the formation of CTC-cluster. Monitoring and administration of IL-2 may benefit pretreated MBC patients and predict disease metastasis.

Citation Format: Zhang Q, Gerratana L, Zhang Y, Flaum L, Shah A, Davis A, Behdad A, Gradishar W, Platanias L, Cristofanilli M. Associations between plasma Interleukin 2 (IL-2) and HER2 expression in circulating tumor cell (CTC) and MYC alterations in circulation tumor DNA (ctDNA) open a new insight on immune microenvironment for patients with metastatic breast cancer (MBC) [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-01-10.

Abstract P4-01-14: Association between interleukin 2 (IL-2) and circulating tumor DNA (ctDNA) is a novel biomarker for patients with metastatic breast cancer (BCa) after systemic therapies

Introduction: The detection and monitoring of ctDNA in metastatic breast cancer showed ability to predict treatment resistance and outcome. But the mechanisms has been a challenge to clinicians. Immune escape and immune tolerance has also been reported to cause BCa progress. Herein, we report a novel finding of the association between plasma IL-2 and the ctDNA in advanced BCa patients who received the systemic therapies, and it is potential utilization in clinic.

Methods: This study enrolled 43 patients with stage III/IV BCa at the Northwestern Memorial Hospital (2016-2017) that had longitudinally detection of ctDNA and circulating tumor cells (CTCs) before (baseline, BL) or 3 months after (first evaluation, FE) systemic therapies respectively. Duplicate whole blood samples (7.5ml/each) were collected in EDTA tubes from these patients. Plasma ctDNA was analyzed using the Guardant360 NGS-based assay (Guardant Health) and CTC enrichment and enumeration were performed in FDA approved semi-automated fluorescence CELLTRACKS ANALYZERII® System (Menarini Silicon Biosystems) by using CELLSEARCH® CXC Kit (Menarini). ELISA (Fisher) for IL-2 was performed by using patients' plasma. Database of IL-2, ctDNA and CTCs was generated and linked with clinical database. Kruskal-Wallis test was used for statistics. We previously reported cut-off of 5.7 was used to dichotomize the prognostic value of ctDNA percentage (%ctDNA) in 2018 ASCO. Matched pairs variations between IL2 levels at BL and at FE were tested through Wilcoxon signed-ranks test. Associations between %ctDNA and IL2 levels were explored through Kruskal-Wallis test. The prognostic impact of IL2 was tested through Cox regression.

Results: CTCs ≥ 5 were found in 23 patients at BL and 21 patients in FE respectively. There were 12 patients that had increase CTCs, and 31 patients with similar or less CTCs FE after systemic therapies. Decreased in CTCs was associated with increased IL-2 (P=0.004).The FE analysis showed that IL-2 dropped significantly in patients with CTC stably ≥5 (from 95.84pg to 79.46pg) after therapies (P<0.001). Furthermore, baselineIL-2 levels were significantly higher in patients with % ctDNA levels ≥5.7 (97.15pg) compared to patients with %ctDNA levels <5.7 (68.64pg) (P=0.0027). No other associations were highlighted in respect to age or number of ctDNA alterations. There was no significant variations between BL and FE levels of IL2 were observed according to BCa subtypes nor in respect to baseline %ctDNA ≥5.7 or CTCs ≥5. Compared with low level of BL IL-2 (<78.3pg) group, high level of BL IL-2 (≥78.3pg) had a significant negative impact on overall survival (OS) (P=0.037) in univariate analysis.

Conclusions: Our findings indicated that aggressive BCa with high level ctDNA mutation are associated with high level of IL-2 and immune response in patients with advanced disease. In addition we confirm a reverse correlation between change of IL-2 and change of CTCs potentially indication of immune escape. In summary, the study shows a dynamic relation between IL-2 level and tumor burden (ctDNA) and immune escape (CTCs) suggesting another potential biomarker to monitor interaction between tumor and immune environment.

Citation Format: Zhang Q, Gerratana L, Zhang Y, Flaum L, Shah A, Davis A, Behdad A, Gradishar W, Platanias L, Cristofanilli M. Association between interleukin 2 (IL-2) and circulating tumor DNA (ctDNA) is a novel biomarker for patients with metastatic breast cancer (BCa) after systemic therapies [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-01-14.

Abstract P5-17-02: Dissecting the biology of inflammatory breast cancer (BC) through cell free DNA and a circulating tumor cells (CTC)-derived signature

Background: The biological characteristics conferring Inflammatory BC's (IBC) distinctive and aggressive clinical features are currently not fully clarified. The aim of this study is to dissect IBC's biology through the integration of DNA and CTC-based circulating biomarkers.

Methods: This study retrospectively analyzed 251 Advanced BC (ABC) patients (pts) treated and longitudinally characterized for CTCs and circulating tumor DNA (ctDNA) at Thomas Jefferson University (Philadephia, PA) and Northwestern University (Chicago, IL). CTCs were enumerated through CellSearch (Menarini Silicon Biosystems), and characterized for HER2 expression using the CellSearch CXC Kit, while ctDNA was analyzed using the Guardant360 NGS assay (Guardant Health) and its percentage (%ctDNA) was classified based on the previously reported cut-off of 5.7% (Gerratana et al 2018). A subset of 117 pts was further characterized for circulating cell-free DNA (ccfDNA) through Qubit® dsDNA HS quantitation Assay (Thermo Fisher Scientific) and quantitative real-time PCR assay for ALU DNA repeats on chromosome 1.Associations between clinical characteristics, CTCs-derived biomarkers and IBC were explored through Fisher's exact test; survival was tested though Cox regression and log-rank test.

Results: Of the total 251 pts, 115 were diagnosed with IBC. Among the 117 patients characterized for ccfDNA, 70 had IBC. Median ccfDNA was 1.59 for IBC (IQR 1.02-3.19) and 2.37 for non-IBC (nIBC) (IQR 1.13-3.52), P=0.27. Consistent results were observed for %ctDNA levels (median value: 2 vs 1.6). The impact on OS of ccfDNA after log transformation was significant for the total population (HR 1.73 95%CI: 1.11-2.69) but not in IBC pts (HR 1.40 95%CI: 0.84-2.34). On the other hand, ctDNA high pts had a significantly worse OS (nIBC: HR 5.34 95%CI: 1.70-18.81 P=0.004; IBC: HR 4.05 95%CI: 1.91-8.58 P< 0.001). In the ctDNA high subgroup no differences in total number of CTCs were observed between IBC and nIBC, while significantly lower CTCs were observed in ctDNA low IBC pts (P=0.0097). The ctDNA low IBC subgroup had a higher incidence of HER2 positive BC (P=0.003) and a significantly lower incidence of CTCs clusters (P=0.006), HER2 positive CTCs (P=0.041). Notably, no associations were observed with stage at baseline, number of metastatic sites, liver, lung and visceral involvement. On the other hand, the ctDNA_high IBC subgroup was characterized by a lower incidence in liver, bone and visceral involvement (P=0.017, P=0.014 and P=0.03 respectively) and a marginally high incidence in soft tissue involvement (0.084). Moreover, IBC diagnosis conferred a significantly worse prognosis only in the ctDNA low subgroup (OS at 12 months nIBC: 100% vs IBC: 70%; P=0.049), while no differences were observed in the ctDNA_high subgroup (OS at 12 months nIBC: 29% vs IBC: 26%; P=0.767).

Conclusion: ctDNA is able to stratify BC according to aggressiveness independently from the sites and type of metastases, both in the IBC and nIBC subgroups. IBC has a distinctive CTCs/ctDNA-based signature, in particular ctDNAlow pts have a lower incidence of HER2 positive CTCs and CTC clusters. This signature is probably due to predominant lymphatic metastatic spread and aggressive phenotype.

Citation Format: Gerratana L, Zhang Q, Wang C, Shah A, Davis AA, Ye Z, Zhang Y, Abu-Khalaf M, Flaum L, Strickland K, Rossi G, Behdad A, Gradishar W, Platanias L, Yang H, Cristofanilli M. Dissecting the biology of inflammatory breast cancer (BC) through cell free DNA and a circulating tumor cells (CTC)-derived signature [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P5-17-02.

Interleukin-2 fusion protein: an investigational therapy for interleukin-2 receptor expressing malignancies

DAB389IL-2 is an interleukin-2 receptor (IL-2R) specific fusion protein with a molecular weight of 58 kD containing the enzymatic and translocation domains of diphtheria toxin (DT) and human IL-2. This fusion protein is able to direct the cytocidal action of the DT enzymatic region only to cells which bear the IL-2R. The human IL-2R exists in three forms: low, intermediate and high affinity. The high-affinity form is believed to be the biologically relevant form on mature, activated T-lymphocytes, B-lymphocytes and monocytes. DAB389IL-2 is able to bind selectively to the high-affinity IL-2R in a concentration-dependent manner, and once bound is internalised via receptor-mediated endocytosis. Upon acidification of the formed vesicle, the enzymatic portion of the fusion protein is believed to pass into the cytosol where it ultimately inhibits protein synthesis by inactivation of elongation factor-2, resulting in cell death. The constitutive expression of the IL-2R on certain leukaemic and lymphomatous cells of T and B cell origin has been reported to occur in patients with chronic lymphocytic leukaemia, cutaneous T cell lymphoma (CTCL), Hodgkin's disease and non-Hodgkin's lymphomas (NHLs). A multicentre DAB389IL-2 dose-escalation study of patients with IL-2R expressing lymphomas has been conducted. A 10-fold range of doses were evaluated on a five-daily dose schedule. Patients received up to six courses, with an additional two courses permitted for patients with partial responses that appeared to be still improving after six courses. Most adverse experiences were transient and mild. Preliminary assessment of response indicated five complete responses (CR, duration ongoing at 20, 11, 7, 5 and 4 months) and seven partial responses (PR, duration 3-20 months) in the 35 patients with CTCL. One CR (duration > 20 months) in a patient with NHL (Lennett's lymphoma) and two PR (duration 9 and 2 months) in 17 patients with B-cell NHL have been observed. Based on the mode of action of DAB389IL-2, its safety profile, and the patient responses associated with the phase I/II clinical trials, a phase III programme in CTCL patients has been initiated and plans for additional trials in NHL patients are targeted for 1996.

Feline parvovirus propagates in cat bone marrow cultures and inhibits hematopoietic colony formation in vitro

Feline parvovirus (FPV) causes leukopenia in naturally infected cats. We investigated the mechanism of hematopoietic depression by this virus in feline bone marrow cultured in vitro. In suspension cultures we demonstrated FPV propagation and replication using DNA molecular hybridization. Viral RNA and DNA were observed by in situ hybridization in about 10% of marrow cells at day 3. Granulocytes and their precursors were virtually absent from infected cultures after six days. Infected cells showed viral capsid protein predominantly in nuclei by immunofluorescence. In clonal assays, FPV most efficiently inhibited hematopoietic colony formation by myeloid progenitor cells (CFU-GM), but erythroid colony formation (BFU-E and CFU-E-derived) was also depressed in the presence of virus. Inhibition of colony formation could be abrogated by physical inactivation of the virus or preincubation with specific neutralizing antibodies. Recombinant human colony stimulating factors GM-CSF and G-CSF supported feline myelopoiesis in progenitor assays, and FPV completely inhibited factor dependent colony formation.

Lymphocyte phenotype and lymphokines following anti-thymocyte globulin therapy in patients with aplastic anaemia

Twenty-two patients with adult onset aplastic anaemia were analysed before and after therapy with anti-thymocyte globulin (ATG). Lymphocyte phenotype, lymphokine levels or production, and haematopoietic progenitor cell number were measured 3 months after therapy; clinical response was determined 1 year post-therapy. By flow cytometry there was a significant reduction in both the proportion and absolute number of peripheral blood lymphocytes expressing activation antigen Tac (IL-2 receptor) and in the proportion of HLA-DR+ lymphocytes. For T cells bearing HLA-DR, there were proportional decreases in both activated helper and suppressor cells. There was no statistically significant difference pre-ATG to post-ATG in the absolute numbers of total, helper and suppressor lymphocytes. In all 10 haematologic responders the number of Tac bearing lymphocytes after ATG therapy was in the normal range, but half of 12 non-responding patients continued to have abnormally elevated numbers of Tac+ T cells. The proportion of Tac+ cells were not related to transfusion history. Gamma-interferon levels in serum by radioimmunoassay were elevated in almost half the aplastic patients; post-ATG, gamma-interferon was detectable in only three patients. Haematologic response to ATG therapy was associated with increased numbers of haematopoietic progenitors post-treatment, but pre-treatment values were not predictive of a response. These results are consistent with a pathogenic role for activated T-cells and their lymphokine products and suggest that the target of ATG therapy may be a Tac+ lymphocyte.

Hematologic effects of immunotherapy with lymphokine-activated killer cells and recombinant interleukin-2 in cancer patients

Immunotherapy with interleukin-2 (IL-2) and lymphokine-activated killer (LAK) cells generated from autologous lymphocytes has produced significant tumor regressions in patients with advanced cancer. In the current study, we reviewed the hematologic effects associated with this therapy in our initial 42 patients. Eighty-eight percent of the treated patients developed anemia that required greater than or equal to 4 units of red cell transfusions, and 43% received at least 8 units. Only a blood loss of 2 to 3 units could be attributed to repeated phlebotomy, cytophereses, and hemodilution. IL-2 administration also resulted in thrombocytopenia as well as lymphopenia and eosinophilia. Forty-three percent of patients developed platelet counts of less than or equal to 50,000/microL, and 36% of the total group required platelet transfusions. Mild neutropenia and a rebound lymphocytosis followed discontinuation of IL-2 treatment. To explore the possible mechanisms for these hematologic effects, standard hematopoietic colony assays were conducted on serial blood samples from five patients. IL-2 produced a significant decline in circulating erythroid (BFU-E) and granulocytic/macrophage (CFU-C) progenitors, which rebounded after the discontinuation of IL-2 therapy. Infusion of IL-2 also resulted in measurable serum levels of gamma-interferon. Some of the hematologic effects of immunotherapy with LAK cells and IL-2 may be the result of IL-2-mediated suppression of hematopoiesis.

Neutropenia associated with large granular lymphocytes responsive to corticosteroids in vitro and in vivo

We describe a patient with neutropenia associated with increased circulating large granular lymphocytes (LGL). Absolute neutropenia was accompanied by the absence of myeloid precursor cells in the bone marrow. No myeloid progenitor cells (CFU-C) could be detected by in vitro colony culture. The peripheral blood was also remarkable for the presence of a population of large granular lymphocytes demonstrable by conventional staining. These cells in flow microfluorometry studies expressed antigens Leu 4 (T-cell antigen receptor), Leu 7 (natural killer cell marker), Leu 2 (suppressor cell marker), and HLA-DR (activation marker); they lacked Leu 1 (a pan-T cell antigen), Leu 3 (helper cell marker) and Tac (interleukin 2 receptor). Hematopoietic colony formation in vitro improved with addition of corticosteroids to the culture medium or elimination of the LGL population with complement-mediated cytotoxicity. Anti-neutrophil antibodies were present prior to and following therapy. Clinically, administration of prednisone resulted in a normalization of the total white blood cell count and absolute polymorphonuclear cell number, an increase into the normal range of the number of CFU-C, and elimination of the LGL population. In this case of steroid-responsive LGL-associated neutropenia, laboratory studies suggested direct suppression of myelopoiesis by steroid-responsive LGL.