- Email: [email protected]
238 Chromosomal instability in cancer and immune-related genes
Abstracts probes to obtain a full picture of the genomic aberrations. SNP arrays are a powerful tool to detect copy number changes down to deletions/ gains/amplifications affecting single genes or parts thereof and in addition provide information on LOH with or without copy number changes. We compared the different SNP array results from single patients obtained from different tissues/sources e.g. tumor vs plasma ctDNA and/or BM-derived disseminated tumor cells (DTCs) vs ctDNA from PB-plasma and ctDNA from BM-plasma. Results: Our data indicate that SNP array analyses allow retrieving the genomic aberrations from 1 ml of cell-free PB- and/or BM-plasma DNA. The detectability of ctDNA in the PB- and BM-plasma depends on the tumor cell infiltration rate in the BM and not on the concentration of cfDNA in the particular plasma sample as our preliminary data show. In the analyzed patient cohort, the majority of detected genomic aberrations were concordant between BM-/PB-plasma DNA and BM-DTCs from the same patient. However, while ctDNAs from BM-plasma and BM-DTCs (n = 3) showed the identical genomic aberrations, ctDNAs from BM- and PB-plasma did not share all genomic aberrations. Conclusions: Plasma-derived ctDNA may serve as an excellent source for tumor genome analysis, especially in cases in which bioptic or surgical interventions carry a high risk for the patient. The findings in this work contradict the current view that all bodily fluids harbor a cross-section of the tumor genome. Therefore, the comparison of datasets obtained from different intra-patient locations will increase the detectability of tumor heterogeneity and thus the understanding of tumor dynamics and progression. No conflict of interest. 236 POSTER Identification and molecular characterization of pure tumor cells from minute formalin-fixed paraffin embedded (FFPE) samples enhanced by image-based microchip sorting technology C. Bolognesi1 , C. Forcato1 , G. Buson1 , F. Fontana1 , G. Signorini2 , G. Medoro2 , H. Morreau3 , M. Barberis4 , W.E. Corver3 , N. Manaresi1 . 1 Silicon Biosystems S.p.A., Biology, Bologna, Italy; 2 Silicon Biosystems S.p.A., Hardware software firmware, Bologna, Italy; 3 Leiden University Medical Center, Pathology, Leiden, Netherlands; 4 European Institute of Oncology, Biology, Milan, Italy Background: We present an innovative workflow combining enzymatic FFPE samples disaggregation and fluorescent staining with DEPArray™ technology to solve the pressing needs in preparation of FFPE samples for genomic analysis: small sample size, unwanted admixture of normal cells, analysis of tumor cell subpopulations present at low percentages. Material and Methods: We disaggregated into cell suspension a total of 23 archival FFPE samples from 1 ovarian, 2 pancreatic, 9 lung, 11 colon cancer patients. The cell mixture was stained for anti-Keratin and anti-Vimentin as well as nuclear stain, that provides an additional cell selection criteria linked to ploidy. We sorted by DEPArray™ precise number (range=5–600) of pure homogenous cells from the major population of tumor cells, the contaminant stromal cells, and other minority tumor cell types indicative of epithelial-to-mesenchymal transition (EMT). After direct lysis of these sorted cells and of unsorted samples, we generated sequencing libraries using IonTorrent AmpliSeq™ Cancer Hotspot Panel v2, and then sequenced with IonTorrent™ PGM. Results: On several loci, we detected somatic mutations with 100% variant frequency, only observable as heterozygous in the unsorted samples and as wild-type in stromal cells of same patient, confirming 100% purity of sorted cells. Moreover, in the rare EMT-phenotype subpopulations we identified clear somatic mutations, shared by tumor prevalent population, confirming their tumor origin. Frequently, for loci harboring germ-line heterozygous SNPs with variant frequency around 50% for pure stromal cells, we readily detected loss-of-heterozygosis in tumor cells subpopulations as binary (0%/100%) variants. Quantitative traits such as copy number gains and losses were also reproducibly identified in tumor cell replicates as deviations from the 50% variant frequency of germline SNPs of pure stromal cells. We observed an excellent coverage uniformity (mean=96%) for recoveries in the range of 81–600 cells. However, mean uniformity gradually decreases for lower cell numbers, probably due to formalin-dependent DNA damage combined with the low amount of DNA material. Nevertheless, with cell numbers of 60 or more, we observed very few outlier false positive variants with frequency higher than 10%, which can thus be assumed as a conservative sensitive and specific threshold for true-positive calling. Conclusions: We demonstrate to address two pressing problems in preparation of FFPE samples for genomic analysis: small sample size and unwanted admixture of normal cells. Analyzing homogenous cell subpopulations boosts signal-to-noise ratio working around inherent
S39 sensitivity/specificity trade-offs of rare-variant calls. Sorting tumor cell subpopulations reveals their genetic characteristics confirming their tumor origin. Conflict of interest: Other Substantive Relationships: Silicon Biosystems S.p.A employee. 237 POSTER The genetic heterogeneity of circulating tumor cells: a longitudinal study in breast cancer patients M. Terracciano1 , G. Buson1 , C. Forcato1 , F. De Luca2 , F. Galardi2 , V. Sero1 , M. Pestrin2 , S. Gabellini2 , C. Koestler3 , Z.T. Czyz3 , G. Bregola1 , P. Tononi1 , C. Bolognesi1 , F. Fontana1 , G. Medoro1 , B. Polzer3 , A. Di Leo2 , C.A. Klein3 , N. Manaresi1 . 1 Silicon Biosystems S.p.A., Biology, Bologna, Italy; 2 Translational Research Laboratory, Prato Hospital, Medical Oncology, Prato, Italy; 3 Fraunhofer ITEM-R, Toxicology and Experimental Medicine, Regensburg, Germany Background: Despite initially successful therapeutic management of breast cancer, a relevant percentage of patients relapse with metastatic disease. The genetic understanding underlying this aspect is still fragmentary but circulating tumor cells (CTCs) represent today a reliable and non-invasive disease monitoring strategy, and offer repeatable access to the detection of tumor genetic heterogeneity. In this study, we performed genetic characterization on pure single CTCs isolated from metastatic breast cancer patients, aiming to provide a full genetic picture of multiple single CTCs for each patient. Methods: Peripheral blood from 4 de novo diagnosed metastatic breast cancer patients, ER+/HER2−, treated either with hormonal or chemotherapy as first line therapy, was collected at three different time points (before start, after one cycle of treatment and at tumor progression). The CTC enrichment and staining was performed with the CellSearch® system and from each sample individual CTCs were sorted with DEPArray™ platform. The DNA of each single CTC was amplified with Ampli1™ WGA kit and Genome Integrity Index (GII) was assessed by Ampli1™ QC kit. WGA output was used for genome-wide single cell copy number variation (CNV) analysis with Agilent SurePrint 180k array for comparative genomic hybridization (aCGH). Furthermore Ampli1™ WGA products were also sequenced with Ampli1™ CHP Custom Panel on the IonTorrent PGM, at 1000x average coverage on Ion 316 or 318 chip. Results: Almost 70% of the collected CTCs showed high GII, as measured by Ampli1™ QC kit (GII 3). For each time point, multiple single CTCs were selected for aCGH and mutation analysis. As expected, aberrant karyotype of CK+/CD45-/DAPI+ CTCs was confirmed by the aCGH analysis. Interestingly, some of the identified CTCs were also CD44+, a cell surface protein associated with metastasis-related phenotypes. The Ampli1™ CHP Custom Panel Beta revealed several COSMIC mutations in coding sequence of different genes, which were further confirmed by the Sanger methodology. All sequence variants for which Sanger sequencing was completed were concordant, in terms of genotype and variant frequency. Conclusions: For each cell a complete genetic picture was created, highlighting for all patients a great genetic heterogeneity among cells showing the same phenotype and also within the same time point. In 3 out of 4 patients aCGH analysis pointed out a global change in terms of genetic aberrations across the different time points. The combination of cell sorting by DEPArray™ platform providing single CTCs with 100% purity, and the robustness of the genetic analysis data following Ampli1™ WGA, provide a promising method to understand CTCs heterogeneity and their clinical implications. Conflict of interest: Other Substantive Relationships: Silicon Biosystems S.p.A. employee. 238 POSTER Chromosomal instability in cancer and immune-related genes S.F. Madden1 , D.M. Collins2 , N. Gaynor2 , W.M. Gallagher1 , N. O’Donovan2 , J. Crown3 . 1 University College Dublin, UCD School of Biomolecular and Biomedical Science, UCD Conway Institute, Dublin, Ireland; 2 Dublin City University, National Institute for Cellular Biotechnology, Dublin, Ireland; 3 St. Vincent’s University Hospital, Dept. of Oncology, Dublin, Ireland Background: The failure of molecularly targeted therapy (MTT) to cure most cancers which contain the relevant target is likely due to intrapatient heterogeneity, the development of bypass signalling pathways and mutational load. An alternative strategy, using “immune checkpoint inhibitors” (ICI) has proven active in several solid tumours, including tumours with de novo or acquired resistance to chemotherapy (CT) and MTT, e.g. melanoma, renal cell and non-small cell lung cancer.
S40
Abstracts
An appreciable minority achieve durable remission. We hypothesise that retained ICI activity in MTT/CT resistant cancers may be partly due to greater genetic stability in immune-related genes compared to non-immune genes in tumours, providing a more inherently stable target. Materials and Methods: Using publicly available array comparative genomic hybridization (aCGH) data for 12 cancer (sub)types: breast (BC)-HER2+, hormone receptor positive (HR+) and triple negative (TN), prostate (PC), melanoma (MEL), ovarian (OV), lung-squamous (LU), colon (CO), mesothelioma (MES), cervical (CSCC), glioblastoma (GBM), chronic lymphocytic leukemia (CLL)) across 554 samples, we compared the frequency of amplification (AMP)/deletion (DEL) of immune-related genes to non-immune genes. Chromosomal regions of AMP/DEL were identified using the R package snapCGH. Genelists were mapped to AMP/DEL chromosome regions as a surrogate for the level of genetic instability within genesets of interest then compared with background conservation levels. Genelists were collated from Gene Ontology, the immunogenicrelated information source, innate DB and Immune Response In Silico (IRIS) datasets and literature searches. Benchmark genelists included oncogenes, tumour suppressors, housekeeping genes and the frequently targeted PI3Kinase and MAPK pathways (KEGG database). All p-values were adjusted using the Benjamini and Hochberg method. Results: IRIS displayed significantly lower frequencies of gene AMP/DEL than all genes included in the aCGH analysis for PC (p = 0.01), OV (p=3.31×10−5 ), MES (p=1.78×10−4 ), BC (HER2+ (p=5.26×10−3 ), HR+ (p = 0.01), TN (p=6.09×10−8 )), LU (p=1.98×10−3 ), CO (p=4.42×10−4 ), CLL (p = 0.01) and across all cancers combined (p=1.97×10−20 ). In contrast, comparison of AMP/DEL frequencies in genes involved in the PI3K signalling pathway (n = 239) showed significantly lower frequencies of AMP/DEL in PC only. In addition, we found that a list of current or potential immune checkpoint genes (n = 28) were relatively conserved in HER2+ BC (p = 0.02), TNBC (p = 0.04), OV (p = 0.03), GBM (p = 0.02), CO (p=4.08×10−6 ) and across all cancers combined (p=2.83×10−7 ). Conclusions: These results support the hypothesis that stability of immune genes might explain the relative activity of ICI in otherwise drug-resistant cancers. Conflict of interest: Corporate-sponsored Research: Dr. Denis Collins is supported by a Roche Postdoctoral Fellowship. 239 POSTER Tyrosine kinase inhibitors (TKIs) enhance response to HER2 targeted antibodies L. Ivers1 , D. Collins1 , A. Canonici1 , G. Gullo2 , N. O’Donovan1 , J. Crown2 . Dublin City University, National Institute for Cellular Biotechnology, Dublin, Ireland; 2 St Vincent’s University Hospital, Medical Oncology, Dublin, Ireland 1
Background: Trastuzumab (T) combined with pertuzumab (P) is approved for metastatic HER2 positive breast cancer. Combined treatment with L and T prior to surgery has been shown to increase the pathological complete response rate. The aim of this study was to examine the potential therapeutic benefit of combining T and P with TKIs, including L and the irreversible panHER TKIs, neratinib (N) and afatinib (A). Materials and Methods: HER2+ cell lines were treated with T, P, (Roche), L, N and A (Sequoia Research Products) as single agents, double combinations (antibody + TKI or 2 antibodies) or triple combinations (T + P + TKI), with/without amphiregulin (5 ng/mL) and heregulin-1b (20 ng/mL). One T-sensitive (SKBR3) and 2 T-resistant cell lines (MDAMB-453, HCC1569) were tested. Results: In SKBR3 cells, the TKIs (L, N, A) enhanced response when combined with T. P alone showed no anti-proliferative activity and did not enhance response to T. P enhanced response to the combination of T and L, whereas P did not enhance response to T with N or A. In the T-resistant cell lines, P+N produced a slight increase in antiproliferative activity whereas T+N did not (Table 1). However, the triple combination (T+P+N) showed enhanced response compared to P and N (MDA-MB-453: p = 0.05; HCC1569: p = 0.04). L or A combined with T and/or P did not enhance compared to L or A alone in the resistant cell lines.
Addition of amphiregulin (AREG) and heregulin-1b (HRG) stimulated proliferation of SKBR3 and MDA-MB-453, but not HCC1569 cells. In the presence of the growth factors, T, P, or T+P, did not inhibit growth of the 3 cell lines, and the TKIs also showed significantly reduced activity. In SKBR3 cells and HCC1569 cells, the triple combinations (T+P+TKI) showed the strongest antiproliferative effect. In MDA-MB-453 cells, T and P plus L showed enhanced response whereas the triple combinations with the irreversible TKIs did not further enhance response compared to T or P plus TKI. Conclusions: Our results support the concept of studying the addition of a TKI to combined anti-HER2 monoclonal antibody therapy in patients with breast cancer. Conflict of interest: Other Substantive Relationships: JC, NOD and DC have received research funding from Roche. JC, AC and NOD have received research funding from Boehringer Ingelheim. JC and NOD have received research funding from GlaxoSmithKline. 240 POSTER Expanding indications to liver surgery by CD34+ autologous cells administration: An overview of 401 consecutive cases at a single institution M. Maestri1 , G. Bocca1 , J. Galafassi1 , G. Lionetto1 , E. Monti1 , G. Costantini1 , N. Murgante1 , A. Peloso1 , S. Gerardini1 , V. Vaccaro1 , S. Brugnatelli2 , A. Pagani2 , L. Moramarco3 , L. Romanini4 , P. Quaretti3 . 1 Fondazione IRCCS Policlinico S. Matteo, Surgery, Pavia, Italy; 2 Fondazione IRCCS Policlinico S. Matteo, Oncology, Pavia, Italy; 3 Fondazione IRCCS Policlinico S. Matteo, Interventional Radiology, Pavia, Italy; 4 Fondazione IRCCS Policlinico S. Matteo, Radiology, Pavia, Italy Background: Liver surgery is regarded as a standard cure for several malignancies. However the amount of parenchyma that can be removed is limited. This study presents the result of liver surgery at a single institution with a special attention to the effect of preliminary administration of multipotent CD34+ cells to boost the regeneration and allow the safe removal of large neoplasms. Material and Methods: 401 pts underwent liver surgery from 2002 to 2015. 18 out of 401 were deemed technically resectable but the perspective remaining parenchyma was below 30%. They were enrolled in a rescue protocol, whose aim was to boost the liver regeneration by conventional portal embolization (group 1, n = 9) or portal embolization plus autologous infusion of CD34+ cells. The whole cohort was evaluated for postoperative complications, margin status, overall survival and time to recurrence. Results: As expected, the number of cancer nodules, margin status and size of lesions have a significant impact on survival (survival analysis, p < 0.05). While single lesions with a R0 margin have the best chance to get cured, even patients with a large tumor burden can experience longterm survival. Patients in the CD 34+ group did not experience a liver recurrence, which is common after the conventional portal embolization (non parametric ANOVA, group 1 vs. group 2 p < 0.05). Complications after liver surgery were similar between pts. treated with conventional surgery vs. pts. enrolled in the rescue protocol (group 1 and 2). Overall perioperative mortality was <1%. Conclusions: Extreme liver surgery is safe and liver regeneration by stem cells seems a promising option, since the results of our study suggest that even those with a large involvement can be rescued to surgery. Compared to conventional regenerative protocols, the administration of CD34+ cells could offer an additional advantage, since none of the treated pts. developed new liver nodules through the follow up. No conflict of interest.
Table 1 (abstract 239). % growth (relative to control) of HER2+ cells treated with T, P, N alone and in combination, with/without AREG and (HRG)
SKRB3 + AREG/HRG MDA-MB-453 + AREG/HRG HCC1569 + AREG/HRG
T
P
N
T+P
T+N
P+N
T+P+N
Medium + AREG/HRG
56.7±2.6 156.4±10.1 114.6±7.9 157.0±1.9 91.3±0.9 88.9±10.2
92.2±8.8 168.2±14.6 119.5±9.4 163.9±5.3 95.1±3.0 105.0±9.3
59.0±5.1 164.2±13.1 89.2±6.3 107.3±14.6 39.6±1.8 96.3±4.7
53.6±4.8 155.5±12.4 111.7±6.3 156.0±7.3 80.4±1.4 93.2±8.3
28.1±2.9 156.3±2.5 88.6±5.8 96.8±10.9 38.2±1.8 87.1±4.5
44.5±9.3 74.3±5.4 83.3±7.2 91.8±8.1 29.5±1.3 82.7±10.7
31.3±4.8 52.0±7.7 71.8±3.2 88.8±9.8 25.4±1.7 61.2±7.8
NA 129.4±4.4 NA 154.9±4.5 NA 94.8±12.3
S39 sensitivity/specificity trade-offs of rare-variant calls. Sorting tumor cell subpopulations reveals their genetic characteristics confirming their tumor origin. Conflict of interest: Other Substantive Relationships: Silicon Biosystems S.p.A employee. 237 POSTER The genetic heterogeneity of circulating tumor cells: a longitudinal study in breast cancer patients M. Terracciano1 , G. Buson1 , C. Forcato1 , F. De Luca2 , F. Galardi2 , V. Sero1 , M. Pestrin2 , S. Gabellini2 , C. Koestler3 , Z.T. Czyz3 , G. Bregola1 , P. Tononi1 , C. Bolognesi1 , F. Fontana1 , G. Medoro1 , B. Polzer3 , A. Di Leo2 , C.A. Klein3 , N. Manaresi1 . 1 Silicon Biosystems S.p.A., Biology, Bologna, Italy; 2 Translational Research Laboratory, Prato Hospital, Medical Oncology, Prato, Italy; 3 Fraunhofer ITEM-R, Toxicology and Experimental Medicine, Regensburg, Germany Background: Despite initially successful therapeutic management of breast cancer, a relevant percentage of patients relapse with metastatic disease. The genetic understanding underlying this aspect is still fragmentary but circulating tumor cells (CTCs) represent today a reliable and non-invasive disease monitoring strategy, and offer repeatable access to the detection of tumor genetic heterogeneity. In this study, we performed genetic characterization on pure single CTCs isolated from metastatic breast cancer patients, aiming to provide a full genetic picture of multiple single CTCs for each patient. Methods: Peripheral blood from 4 de novo diagnosed metastatic breast cancer patients, ER+/HER2−, treated either with hormonal or chemotherapy as first line therapy, was collected at three different time points (before start, after one cycle of treatment and at tumor progression). The CTC enrichment and staining was performed with the CellSearch® system and from each sample individual CTCs were sorted with DEPArray™ platform. The DNA of each single CTC was amplified with Ampli1™ WGA kit and Genome Integrity Index (GII) was assessed by Ampli1™ QC kit. WGA output was used for genome-wide single cell copy number variation (CNV) analysis with Agilent SurePrint 180k array for comparative genomic hybridization (aCGH). Furthermore Ampli1™ WGA products were also sequenced with Ampli1™ CHP Custom Panel on the IonTorrent PGM, at 1000x average coverage on Ion 316 or 318 chip. Results: Almost 70% of the collected CTCs showed high GII, as measured by Ampli1™ QC kit (GII 3). For each time point, multiple single CTCs were selected for aCGH and mutation analysis. As expected, aberrant karyotype of CK+/CD45-/DAPI+ CTCs was confirmed by the aCGH analysis. Interestingly, some of the identified CTCs were also CD44+, a cell surface protein associated with metastasis-related phenotypes. The Ampli1™ CHP Custom Panel Beta revealed several COSMIC mutations in coding sequence of different genes, which were further confirmed by the Sanger methodology. All sequence variants for which Sanger sequencing was completed were concordant, in terms of genotype and variant frequency. Conclusions: For each cell a complete genetic picture was created, highlighting for all patients a great genetic heterogeneity among cells showing the same phenotype and also within the same time point. In 3 out of 4 patients aCGH analysis pointed out a global change in terms of genetic aberrations across the different time points. The combination of cell sorting by DEPArray™ platform providing single CTCs with 100% purity, and the robustness of the genetic analysis data following Ampli1™ WGA, provide a promising method to understand CTCs heterogeneity and their clinical implications. Conflict of interest: Other Substantive Relationships: Silicon Biosystems S.p.A. employee. 238 POSTER Chromosomal instability in cancer and immune-related genes S.F. Madden1 , D.M. Collins2 , N. Gaynor2 , W.M. Gallagher1 , N. O’Donovan2 , J. Crown3 . 1 University College Dublin, UCD School of Biomolecular and Biomedical Science, UCD Conway Institute, Dublin, Ireland; 2 Dublin City University, National Institute for Cellular Biotechnology, Dublin, Ireland; 3 St. Vincent’s University Hospital, Dept. of Oncology, Dublin, Ireland Background: The failure of molecularly targeted therapy (MTT) to cure most cancers which contain the relevant target is likely due to intrapatient heterogeneity, the development of bypass signalling pathways and mutational load. An alternative strategy, using “immune checkpoint inhibitors” (ICI) has proven active in several solid tumours, including tumours with de novo or acquired resistance to chemotherapy (CT) and MTT, e.g. melanoma, renal cell and non-small cell lung cancer.
S40
Abstracts
An appreciable minority achieve durable remission. We hypothesise that retained ICI activity in MTT/CT resistant cancers may be partly due to greater genetic stability in immune-related genes compared to non-immune genes in tumours, providing a more inherently stable target. Materials and Methods: Using publicly available array comparative genomic hybridization (aCGH) data for 12 cancer (sub)types: breast (BC)-HER2+, hormone receptor positive (HR+) and triple negative (TN), prostate (PC), melanoma (MEL), ovarian (OV), lung-squamous (LU), colon (CO), mesothelioma (MES), cervical (CSCC), glioblastoma (GBM), chronic lymphocytic leukemia (CLL)) across 554 samples, we compared the frequency of amplification (AMP)/deletion (DEL) of immune-related genes to non-immune genes. Chromosomal regions of AMP/DEL were identified using the R package snapCGH. Genelists were mapped to AMP/DEL chromosome regions as a surrogate for the level of genetic instability within genesets of interest then compared with background conservation levels. Genelists were collated from Gene Ontology, the immunogenicrelated information source, innate DB and Immune Response In Silico (IRIS) datasets and literature searches. Benchmark genelists included oncogenes, tumour suppressors, housekeeping genes and the frequently targeted PI3Kinase and MAPK pathways (KEGG database). All p-values were adjusted using the Benjamini and Hochberg method. Results: IRIS displayed significantly lower frequencies of gene AMP/DEL than all genes included in the aCGH analysis for PC (p = 0.01), OV (p=3.31×10−5 ), MES (p=1.78×10−4 ), BC (HER2+ (p=5.26×10−3 ), HR+ (p = 0.01), TN (p=6.09×10−8 )), LU (p=1.98×10−3 ), CO (p=4.42×10−4 ), CLL (p = 0.01) and across all cancers combined (p=1.97×10−20 ). In contrast, comparison of AMP/DEL frequencies in genes involved in the PI3K signalling pathway (n = 239) showed significantly lower frequencies of AMP/DEL in PC only. In addition, we found that a list of current or potential immune checkpoint genes (n = 28) were relatively conserved in HER2+ BC (p = 0.02), TNBC (p = 0.04), OV (p = 0.03), GBM (p = 0.02), CO (p=4.08×10−6 ) and across all cancers combined (p=2.83×10−7 ). Conclusions: These results support the hypothesis that stability of immune genes might explain the relative activity of ICI in otherwise drug-resistant cancers. Conflict of interest: Corporate-sponsored Research: Dr. Denis Collins is supported by a Roche Postdoctoral Fellowship. 239 POSTER Tyrosine kinase inhibitors (TKIs) enhance response to HER2 targeted antibodies L. Ivers1 , D. Collins1 , A. Canonici1 , G. Gullo2 , N. O’Donovan1 , J. Crown2 . Dublin City University, National Institute for Cellular Biotechnology, Dublin, Ireland; 2 St Vincent’s University Hospital, Medical Oncology, Dublin, Ireland 1
Background: Trastuzumab (T) combined with pertuzumab (P) is approved for metastatic HER2 positive breast cancer. Combined treatment with L and T prior to surgery has been shown to increase the pathological complete response rate. The aim of this study was to examine the potential therapeutic benefit of combining T and P with TKIs, including L and the irreversible panHER TKIs, neratinib (N) and afatinib (A). Materials and Methods: HER2+ cell lines were treated with T, P, (Roche), L, N and A (Sequoia Research Products) as single agents, double combinations (antibody + TKI or 2 antibodies) or triple combinations (T + P + TKI), with/without amphiregulin (5 ng/mL) and heregulin-1b (20 ng/mL). One T-sensitive (SKBR3) and 2 T-resistant cell lines (MDAMB-453, HCC1569) were tested. Results: In SKBR3 cells, the TKIs (L, N, A) enhanced response when combined with T. P alone showed no anti-proliferative activity and did not enhance response to T. P enhanced response to the combination of T and L, whereas P did not enhance response to T with N or A. In the T-resistant cell lines, P+N produced a slight increase in antiproliferative activity whereas T+N did not (Table 1). However, the triple combination (T+P+N) showed enhanced response compared to P and N (MDA-MB-453: p = 0.05; HCC1569: p = 0.04). L or A combined with T and/or P did not enhance compared to L or A alone in the resistant cell lines.
Addition of amphiregulin (AREG) and heregulin-1b (HRG) stimulated proliferation of SKBR3 and MDA-MB-453, but not HCC1569 cells. In the presence of the growth factors, T, P, or T+P, did not inhibit growth of the 3 cell lines, and the TKIs also showed significantly reduced activity. In SKBR3 cells and HCC1569 cells, the triple combinations (T+P+TKI) showed the strongest antiproliferative effect. In MDA-MB-453 cells, T and P plus L showed enhanced response whereas the triple combinations with the irreversible TKIs did not further enhance response compared to T or P plus TKI. Conclusions: Our results support the concept of studying the addition of a TKI to combined anti-HER2 monoclonal antibody therapy in patients with breast cancer. Conflict of interest: Other Substantive Relationships: JC, NOD and DC have received research funding from Roche. JC, AC and NOD have received research funding from Boehringer Ingelheim. JC and NOD have received research funding from GlaxoSmithKline. 240 POSTER Expanding indications to liver surgery by CD34+ autologous cells administration: An overview of 401 consecutive cases at a single institution M. Maestri1 , G. Bocca1 , J. Galafassi1 , G. Lionetto1 , E. Monti1 , G. Costantini1 , N. Murgante1 , A. Peloso1 , S. Gerardini1 , V. Vaccaro1 , S. Brugnatelli2 , A. Pagani2 , L. Moramarco3 , L. Romanini4 , P. Quaretti3 . 1 Fondazione IRCCS Policlinico S. Matteo, Surgery, Pavia, Italy; 2 Fondazione IRCCS Policlinico S. Matteo, Oncology, Pavia, Italy; 3 Fondazione IRCCS Policlinico S. Matteo, Interventional Radiology, Pavia, Italy; 4 Fondazione IRCCS Policlinico S. Matteo, Radiology, Pavia, Italy Background: Liver surgery is regarded as a standard cure for several malignancies. However the amount of parenchyma that can be removed is limited. This study presents the result of liver surgery at a single institution with a special attention to the effect of preliminary administration of multipotent CD34+ cells to boost the regeneration and allow the safe removal of large neoplasms. Material and Methods: 401 pts underwent liver surgery from 2002 to 2015. 18 out of 401 were deemed technically resectable but the perspective remaining parenchyma was below 30%. They were enrolled in a rescue protocol, whose aim was to boost the liver regeneration by conventional portal embolization (group 1, n = 9) or portal embolization plus autologous infusion of CD34+ cells. The whole cohort was evaluated for postoperative complications, margin status, overall survival and time to recurrence. Results: As expected, the number of cancer nodules, margin status and size of lesions have a significant impact on survival (survival analysis, p < 0.05). While single lesions with a R0 margin have the best chance to get cured, even patients with a large tumor burden can experience longterm survival. Patients in the CD 34+ group did not experience a liver recurrence, which is common after the conventional portal embolization (non parametric ANOVA, group 1 vs. group 2 p < 0.05). Complications after liver surgery were similar between pts. treated with conventional surgery vs. pts. enrolled in the rescue protocol (group 1 and 2). Overall perioperative mortality was <1%. Conclusions: Extreme liver surgery is safe and liver regeneration by stem cells seems a promising option, since the results of our study suggest that even those with a large involvement can be rescued to surgery. Compared to conventional regenerative protocols, the administration of CD34+ cells could offer an additional advantage, since none of the treated pts. developed new liver nodules through the follow up. No conflict of interest.
Table 1 (abstract 239). % growth (relative to control) of HER2+ cells treated with T, P, N alone and in combination, with/without AREG and (HRG)
SKRB3 + AREG/HRG MDA-MB-453 + AREG/HRG HCC1569 + AREG/HRG
T
P
N
T+P
T+N
P+N
T+P+N
Medium + AREG/HRG
56.7±2.6 156.4±10.1 114.6±7.9 157.0±1.9 91.3±0.9 88.9±10.2
92.2±8.8 168.2±14.6 119.5±9.4 163.9±5.3 95.1±3.0 105.0±9.3
59.0±5.1 164.2±13.1 89.2±6.3 107.3±14.6 39.6±1.8 96.3±4.7
53.6±4.8 155.5±12.4 111.7±6.3 156.0±7.3 80.4±1.4 93.2±8.3
28.1±2.9 156.3±2.5 88.6±5.8 96.8±10.9 38.2±1.8 87.1±4.5
44.5±9.3 74.3±5.4 83.3±7.2 91.8±8.1 29.5±1.3 82.7±10.7
31.3±4.8 52.0±7.7 71.8±3.2 88.8±9.8 25.4±1.7 61.2±7.8
NA 129.4±4.4 NA 154.9±4.5 NA 94.8±12.3