06.08 Compliance of Endocrine Therapy and Outcomes Following Lumpectomy Without Adjuvant Radiation

Posted on January 25, 2017

M. G. Shah-Khan1, L. Woldanski1, A. L. Kong1  1Medical College Of Wisconsin,Surgical Oncology,Milwaukee, WI, USA

Introduction:

Patient adherence to prescribed endocrine therapy for breast cancer is variable in both the preventative and adjuvant setting.  For patients undergoing breast conserving surgery (BCS), there is data to support forgoing radiation therapy in older women with estrogen receptor positive (ER+) disease.  We sought to determine the rate of initiation and adherence to endocrine therapy in patients undergoing lumpectomy without radiation therapy, and determine the rate of recurrence in these patients.

Methods:

A retrospective review was performed to identify patients treated within our institution from 2000-2014.   Patients included were aged 60 and older and underwent BCS for ER+ invasive cancer (T1-T2) or DCIS, and did not undergo adjuvant radiation therapy.   Chart review was performed to determine the rate of initiation and subsequent adherence to endocrine therapy.   Recurrence and survival information was also evaluated.

Results:

A total of 206 patients were evaluated, with mean age of 76.  Mean follow up was 49 months.  One hundred forty three patients underwent BCS for invasive cancer, and 63 patients for DCIS.  Of the patients with invasive cancer, 89% (n=127) were T1, and 11% (n=16) were T2.  Seventy percent were lymph node negative, 7% were lymph node positive, and 23% had no lymph nodes evaluated.  Sixty eight percent (n=97) initiated endocrine therapy with either tamoxifen or an aromatase inhibitor.  Forty nine patients (30%) adhered to treatment for greater than 2 years, with 19% completing at least 4 years of treatment.   A total of 14 (10%) of patients had a recurrence in the study period.  Twelve patients (8%) had an ipsilateral locoregional recurrence, and 1 patient had a distant recurrence.  One patient developed a contralateral primary invasive breast cancer during follow up.  Twenty six percent of patients (n=37) died during the follow up period, with one death related to metastatic breast cancer.

Patients undergoing lumpectomy for DCIS had a lower rate of initiation of therapy, however rate of compliance over time was not significantly different. Only 35% of DCIS patients (n=22) initiated endocrine therapy, compared to 68% of patients with invasive cancer (p<0.0001).  Twenty percent of patients (n=13) completed at least 2 years of therapy.  Thirteen percent (n=8) of patients completed at least 4 years of therapy.  Four patients (6%) developed a recurrence, 2 with ipsilateral DCIS, and 2 with contralateral primary breast cancers (1 DCIS, 1 invasive).  Three patients (4.7%) died during the study period, with no breast cancer related deaths. 

Conclusion:

In older patients undergoing BCS for T1-T2 ER+ in situ or invasive breast cancer who forego adjuvant radiation, initiation of endocrine therapy is higher in patients with invasive cancer than with DCIS, however compliance decreases in both groups over time.  A low incidence of recurrence and breast cancer related mortality was seen in both groups of patients. 

06.07 The Prognostic Significance of Adrenocortical Carcinomas Identified Incidentally

Posted on January 25, 2017

K. K. Rossfeld1, N. Saunders2, T. Tran13, Q. Duh9, J. Mansour8, S. Maithel2, C. Solorzano10, T. Wang4, J. Glenn4, E. Levine11, S. Weber6, A. Salem6, R. Fields5, G. Poultsides13, T. Pawlik1, J. Phay1, L. A. Shirley1  1Ohio State University,Department Of Surgery,Columbus, OH, USA 2Emory University School Of Medicine,Department Of Surgery,Atlanta, GA, USA 3Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA 4Medical College Of Wisconsin,Department Of Surgery,Milwaukee, WI, USA 5Washington University,Department Of Surgery,St. Louis, MO, USA 6University Of Wisconsin,Department Of Surgery,Madison, WI, USA 7University Of California – San Diego,Department Of Surgery,San Diego, CA, USA 8University Of Texas Southwestern Medical Center,Department Of Surgery,Dallas, TX, USA 9University Of California – San Francisco,Department Of Surgery,San Francisco, CA, USA 10Vanderbilt University Medical Center,Department Of Surgery,Nashville, TN, USA 11Wake Forest University School Of Medicine,Department Of Surgery,Winston-Salem, NC, USA 12New York University School Of Medicine,Department Of Surgery,New York, NY, USA 13Stanford University,Department Of Surgery,Palo Alto, CA, USA

Introduction:
Adrenal tumors, including adrenocortical carcinomas (ACC), are often identified incidentally on imaging, being diagnosed as an “incidentaloma”.   Little is known regarding the difference in prognosis among patients who have an incidental ACC versus those patients with ACC who present with signs or symptoms.

Methods:
Patients who underwent resection of ACC at one of 13 institutions between January 1993 and December 2014 were identified.  ACC tumors were categorized as incidentalomas versus symptomatic adrenal tumors.  Patients without data fields needed for the current study were excluded.  Data on patient age, gender, race, body mass index (BMI), tumor size, tumor weight, presence of metastases, and margin status were collected and analyzed according to incidental versus non-incidental ACC category.  Clinical variables were analyzed using Chi-square, Fisher Exact, or independent T-test as appropriate.  Kaplan-Meier curves for overall (OS) and recurrence-free survival (RFS) were analyzed using log-rank testing.  Cox regression was utilized to analyze potential differences in OS and RFS among relevant clinical variables.

Results:
Among 232 evaluable patients with ACC, 100 (43.1%) were diagnosed with an incidental ACC while 132 (56.9%) patients presented with symptoms/signs.  Clinical and pathological features including age, gender, race, BMI, nodal status, tumor size, tumor weight, and surgical margin status were all comparable among incidental versus non-incidental ACC patients (all p>0.05). However, patients with incidentalomas were more likely to have a T1/T2 tumor (55.8% vs. 33.9%, p<0.01) and were less likely to have metastatic disease at presentation (13.1 % vs. 23.6%, p=0.031).  Patients with an incidental ACC had a better median RFS (29.4 months) compared with patients with a non-incidental ACC (12.5 months, p=0.020).  The improved RFS among patients with incidental ACC tumors persisted when RFS was stratified by the presence of metastasis (p=0.034), but not by T stage (p= 0.29).  Among patients who initially presented with an incidentaloma, those who had tumors that were later noted to be hormonally active on further work-up had a shorter RFS compared with patients with non-functional tumors (8.8 vs. 47.9 months, p=0.002). 

Conclusion:
Patients with resected ACC who had tumors that were identified incidentally had an improved RFS compared with patients who presented with symptoms or signs.  This difference in RFS was likely related to a lower T-stage on presentation.  Patients with incidentalomas that were hormonally active had a shorter RFS than those with non-functional incidentalomas, highlighting potential prognostic importance of evaluating hormonal secretion in asymptomatic patients.
 

06.06 Residual Lymph Node Metastasis After Initial Surgical Treatment Of Differentiated Thyroid Cancer

Posted on January 25, 2017

O. Brown1, J. Miller1, N. Al-Attar1, G. Shaughness1, N. Rosculet1, A. Avram1, D. Hughes1  1University Of Michigan,Ann Arbor, MI, USA

Introduction:  Differentiated thyroid cancer (DTC) is treated with thyroidectomy and removal of malignant lymph nodes identified preoperatively. Despite attempts to resect regional metastatic disease, residual malignant lymph nodes are often found on diagnostic and/or post-therapy radioiodine scans at the time of radioactive iodine treatment. The objective of this study is to identify the most common locations of and reasons for residual lymph node metastases following initial surgical treatment of DTC.

Methods:  In this retrospective study, we reviewed the charts of 352 patients with high-risk DTC treated at the University of Michigan from 2007 to 2014. Postoperatively, all patients underwent pre-ablation 131-Iodine scans with SPECT/CT. The laterality of the primary tumor and the presence and location of nodal metastases were noted. We compared the location of the nodal metastases with cervical lymph node levels that were dissected to determine the reason for residual nodal metastases (incomplete node dissection vs. omission of node dissection in that compartment due to failure to identify nodal involvement preoperatively).

Results: A total of 146/352 (41.48%) patients with high-risk DTC were found to have residual lymph node metastases on diagnostic 131-I SPECT/CT following initial surgical treatment. These 146 patients had a combined total of 218 nodal metastases. Relative to the primary tumor, 71.56% (n=156) were ipsilateral, 22.02% (n=48) were contralateral, and 6.42% (n=14) were undefined, meaning they were classified as central neck (level VI/VII) without a designation of laterality. Cervical lymph node levels VI, III, and II were found to have the greatest frequencies of residual metastases, with 33.94% (n=74), 22.94% (n=50), 18.81% (n=41), respectively. Levels I and V had the lowest frequencies of metastases, with 0.46% (n=1) and 0.00% (n=0), respectively. Residual metastases occurred because of an incomplete lymph node dissection in 72 patients (49.32%), a missed diagnosis in 55 patients (37.67%), or a combination of the two in 19 patients (13.01%).

Conclusion: The surgical management of high-risk DTC should include complete nodal dissection in all involved compartments. Particular attention should be given to complete dissection in levels VI, III, and II, especially ipsilateral to the primary tumor, as these are prevalent areas of residual nodal metastasis. Efforts to improve pre-operative diagnostic techniques and to refine lymph node dissection practices may reduce the incidence of residual nodal metastasis.

06.05 Cancer Independently Worsens and Increases Variability of Short-term Surgical Outcomes

Posted on January 25, 2017

I. L. Leeds1, J. E. Efron1, N. Ahuja1, E. R. Haut1, F. M. Johnston1  1Johns Hopkins University School Of Medicine,Department Of Surgery,Baltimore, MD, USA

Introduction:
There is a growing consensus that common surgical prediction models do not adequately predict surgical outcomes for cancer surgery patients. The purpose of this study was to examine the comparative performance of these models in benign versus malignant gastrointestinal surgery populations to assess their utility as a risk-counseling tool for providers and patients.

Methods:
The American College of Surgeons’ National Surgical Quality Improvement Program’s (NSQIP) 2014 participant use file was queried for patients undergoing elective surgery with a primary procedure code of a major colon, pancreas, or stomach resection. All patients with a diagnosis of gastrointestinal cancer were compared against those without using Chi-square and Wilcox-Mann-Whitney tests. Multivariable logistic regression was performed to identify independent predictors of mortality and morbidity.

Results:
We identified 37,809 cases (16,322 malignant, 21,487 benign; 30,789 colectomies, 5,963 pancreatectomies, 1,057 gastrectomies). All of the following reported differences were statistically significant. The gastrointestinal cancer surgery population was disproportionately male, older than 65, non-white, and less functionally independent. Major comorbidities more prevalent in the non-cancer surgery group included chronic steroid use, preoperative sepsis, and active smoking. Comorbidities more common in the cancer surgery population included disseminated malignancy, diabetes, hypertension, dyspnea, and COPD.

Cancer surgery patients had a longer length of stay (+0.9 days), higher mortality rate (0.9% vs. 0.5%), higher complication rate (26.9% vs. 21.2%), higher total number of complications per surgery (0.5 vs. 0.3), broader variation in total number of complications (σ2= 1.0 vs. 0.8), higher readmission rate (11.4% versus 10.6%), higher pneumonia incidence (2.5% vs. 1.9%), higher reintubation rate (2.1% versus 1.3%), extended ventilator-days (1.6% vs. 1.1%), more kidney injuries (1.4% vs. 0.9%), more in-hospital cardiac arrests (0.7% vs. 0.4%), more blood transfusions (12.0% vs. 6.4%), and more blood clots (1.4% vs. 1.1%).

NSQIP probability variables performed similarly for complications in cancer and non-cancer patients (r= 0.32 vs. 0.31) but underperformed for mortality (r= 0.24 vs. 0.14). Multivariable regression controlling for predictors and procedure type demonstrated that a diagnosis of cancer requiring surgery independently confers an increased risk of death of 34.4% (p=0.034) and 12.0% (p<0.001) increased risk of additional complications as compared to patients with benign disease.

Conclusion:
NSQIP prediction models less effectively evaluate the risk of death in cancer patients as compared to patients undergoing similar resections for benign disease. Furthermore, a diagnosis of cancer independently increases the risk of complication and death from surgical resection. Additional counseling of short-term outcomes uncertainty may be needed.
 

06.04 Oncologic Outcomes Associated with Robotic Pancreatic Resections

Posted on January 25, 2017

P. Briceno1, R. Shridhar3, J. Huston1, S. Kucera2, K. Meredith1  1Florida State University College Of Medicine/Sarasota Memorial Health Care System,Gastrointestinal Oncology,Sarasota, FL, USA 2Florida State University College Of Medicine,Endoscopic Oncology,Sarasota, FL, USA 3University Of Central Florida,Radiation Oncology,Orlando, FL, USA

Introduction:  Surgery remains a key component of treatment for resectable pancreatic adenocarcinoma. Mortality rates have dropped dramatically over the past several decades with improvements in preoperative care, intraoperative surgical techniques and instrumentation, as well as post-operative care. However, improving outcomes from pancreatic resection through minimally invasive surgical approaches has gained interest. While there is widespread adoption of both laparoscopic and robotic resections for cancers of the left pancreas the demanding technical requirements of performing a minimally invasive pancreaticoduodenectomy (RPD) and prolonged learning curve have proven to difficult for most. There are several reports of operative outcomes for patients undergoing robotic pancreatic resections, however the survival in these patients has not been investigated. 

Methods:  Utilizing the National Cancer Database we identified patients who underwent pancreatic resection. We then stratified by operative approach. Outcomes of patients were then compared between open(OP), robotic(RP) and laparoscopic(LP) approaches.  Baseline univariate comparisons of patient characteristics were made for continuous variables using both the Mann-Whitney U and Kruskal Wallis tests as appropriate. Pearson’s Chi-square test was used to compare categorical variables. Survival was evaluated on the basis of time from date of diagnosis to date of death or censoring. Unadjusted survival analyses were performed using the Kaplan-Meier method comparing survival curves with the log-rank test. All statistical tests were two-sided and α (type I) error <0.05 was considered statistically significant. 

Results: We identified N = 17,738 patients who underwent pancreatic resection with a median age of 66 (18-90) years.  There were 14,568 open, 2,739 laparoscopic, and 431 robotic. There were 7943 open whipples (OW) and 3455 open distal pancreatectomies (ODP).  Within the minimally invasive approaches there were 1123 (41%) laparoscopic whipples (LW), 1616 (59%) laparoscopic distal pancreatectomies (LDP), 122 (28.3%) robotic whipple (RW), and 309 (71.7%) robotic distal pancreatectomies (RDP). The 30-day mortality between OP, LP and RP was 3.5%, 2.8%, and 2.6%, p=0.1 however the 90-day mortality was 7.2%, 5.8%, and 4.6%, p=0.004. R0 resections were performed in 78.5% OP, 83.7% LP, and 86.3% RP, p<0.001. The median survival was 30.1 months OP, 37.7 months LP, and 52.3 months RP, p<0.001. However, RP and LP had higher T1 and T2 as well as node negative patients p<0.001 and p<0.001 respectively.  Conversions to open occurred in 24.2% LP and 13.9% RP, p<0.001. Within the LDP and RDP there were 351 (21.7%) and 36 (11.7%) conversions, p<0.001. The LW and RW demonstrated 308 (27.4%) and 29 (23.8%) conversions, p=0.4. 

Conclusion: Our analysis indicates that robotic assisted pancreatic resections are at least equivalent in oncologic outcomes to standard open approaches.  Robotic techniques resulted in fewer conversions to open compared to laparoscopic approaches particularly in the distal pancreatectomy cohorts. 

 

06.03 Gene-expression Classifier Testing in the Management of Cytologically Indeterminate Thyroid Nodules.

Posted on January 25, 2017

M. Affi1, B. Javorsky1, T. Yen1, K. Doffek1, B. Hunt1, P. Tolat1, T. Carroll1, T. Giorgadze1, A. Carr1, D. Evans1, T. Wang1  1Medical College Of Wisconsin,Milwaukee, WI, USA

Introduction:  Up to 30% of thyroid nodules evaluated by fine needle aspiration (FNA) are indeterminate on cytology. Use of a 167 gene mRNA-based gene-expression classifier (GEC) test has been shown to have a high negative predictive value for benign nodules and a 40% malignancy rate for GEC-suspicious nodules. The aim of this study was to determine the rate of malignancy in thyroid nodules indeterminate on initial cytology that underwent repeat FNA in anticipation of GEC testing. 

Methods:  This is a retrospective review of all patients with thyroid nodules with indeterminate cytology ('atypia/follicular cells of unknown significance’ or ‘suspicious for follicular neoplasm’) and who underwent GEC testing between 10/12-5/16. Prior to GEC testing, repeat FNA was performed and cytology results were reviewed by an independent cytologist; only indeterminate samples underwent GEC testing. Patients with GEC-benign nodules were recommended for surveillance. Patients with re-classified cytology that was 'suspicious for malignancy' or GEC-suspicious nodules were referred for thyroidectomy. The rate of clinically significant malignancy (>1cm) in patients undergoing thyroidectomy was determined, based on both initial cytology and results of GEC testing.

Results: 89 patients met inclusion criteria. Independent review of cytology yielded 35 (39%) benign, 48 (54%) indeterminate, 2 (2%) suspicious for malignancy, and 4 (5%) non-diagnostic samples. Of the 35 patients with benign cytology, 2 (6%) underwent thyroidectomy with benign final histopathology. The 2 patients with suspicious cytology had a clinically significant papillary thyroid cancer on final histopathology. Of the remaining 52 samples, 46 had sufficient mRNA for GEC analysis; 17 (37%) were GEC-benign and 29 (63%) were GEC-suspicious. None of the patients with GEC-benign lesions have had surgery; to date, 6 (35%) patients have had no changes on follow-up ultrasound. Of 26 patients with GEC-suspicious samples who underwent thyroidectomy, 18 (69%) had benign pathology, 5 (19%) had a clinically significant malignancy, and 3 (12%) patients had an incidental malignancy. Overall, 32 (36%) patients underwent thyroidectomy; based on the original institutional cytological diagnosis of ‘indeterminate,’ the rate of clinically significant malignancy was 22% (7/32).  

Conclusion: Of the 89 patients with indeterminate initial cytology, repeat FNA with cytological review and GEC testing altered the management of 52 (58%) patients. The clinically significant malignancy rate for GEC-suspicious thyroid nodules was 19%, much lower than the reported 40% rate and equivalent to our institutional rate of malignancy in patients with indeterminate nodules who underwent surgery. Further follow-up of nonsurgical patients with GEC-benign nodules is required to determine the true benefit of GEC testing in patients with cytologically indeterminate thyroid nodules at our institution.

 

06.02 Novel MicroRNA Expression Patterns of Breast Cancer to Predict Survival

Posted on January 25, 2017

T. Kawaguchi1, L. Yan2, Q. Qi2, S. Liu2, K. Takabe1  1Roswell Park Cancer Institute,Breast Surgery, Department Of Surgical Oncology,,Buffalo, NY, USA 2Roswell Park Cancer Institute,Department Of Biostatistics & Bioinformatics,Buffalo, NY, USA

Introduction:
MicroRNAs (miRNAs) are small non-coding RNAs that exert its functions by regulating expression of their target genes. Dysregulations of miRNAs are related with breast cancer (BrCa).The purpose of this study was to classify BrCa with miRNAs expression patterns to predict survival utilizing The Cancer Genome Atlas (TCGA).

Methods:
Both clinical and miRNA-seq data enrolled in TCGA dataset were retrieved from the GDC data portal for analyses, and were evaluate by hierarchical clustering based on bioinformatics analysis. We also evaluate clinical relevance including prognostic analysis based on the novel subclasses using the Cox proportional hazard model. 

Results:
Of 1097 cases enrolled in TCGA dataset, 1052 caseswere used for miRNAs expression data and survival analysis. We devided the cases into “short” (died within 3 years after diagnosis), “long” (lived longer than 5 years), and the others. We identified that 15 miRNAs (let-7a-1, miR-106a, miR-17, miR-184, miR-18a, miR-193a, miR-19a, miR-20a, miR-20b, miR-362, miR-4661, miR-500a, miR-766, miR-92a-1, miR-93) were significantly differently expressed between the long and short group. With the expression pattern of these 15 miRNA, the patients were classified into three clusters. Of the 15 miRNAs, we conducted additional feature selection in a multivariate Cox proportional hazard model, and three miRNAs remain after model selection (miR-106a, miR-766, and miR-93). We generated a risk scoring model with the expression of the three miRNAs based on Cox proportional hazard model. We found that the patients with the high score significantly associated with poor outcome. 

Conclusion:
We demonstrated a novel concept that microRNA expression patterns of BrCa can predict worse survival.
 

06.01 Hospital Variation in Neoadjuvant Therapy for T1-2 Pancreatic Cancer: Short and Long-term Outcomes

Posted on January 25, 2017

M. A. Healy1, S. Shubeck1, W. Burns1, T. L. Frankel1, H. Nathan1  1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA

Introduction:  The benefit of neoadjuvant therapy for resectable pancreatic cancer (PanCa) is controversial, and its use varies significantly between centers. It is unknown whether neoadjuvant therapy practices are associated with short- and long-term patient outcomes across institutions. We sought to evaluate whether institutions that routinely use neoadjuvant therapy achieve superior aggregate outcomes for PanCa patients. 

Methods:  Using the National Cancer Data Base Participant User File, we identified patients with non-metastatic T1-2 PanCa diagnosed from 2006-2013. Hospitals were stratified into quintiles based on risk- and reliability-adjusted rates of neoadjuvant therapy use in resected patients (including chemotherapy and/or radiation). Post-operative and survival outcomes were compared between hospitals using neoadjuvant therapy frequently (Neo-Hi) vs infrequently (Neo-Lo).

Results: We identified 20,646 patients with median age 72 years and 25% T1 tumors. Among 815 treating hospitals, adjusted rates of neoadjuvant therapy varied almost 30-fold (2.9%-83%). Patients at Neo-Hi vs Neo-Lo hospitals had higher readmission rates (8.8% vs 7.4%; P<0.01) but similar 30-day mortality (6.0% vs 5.8%;P=0.6). Neo-Hi vs Neo-Lo hospitals had lower surgical margin positivity (17% vs 21%; P<0.001) and node positivity (56% vs 63%; P<0.001). Resected patients at Neo-Hi vs Neo-Lo hospitals had marginally improved adjusted overall survival at 2 years (48% vs 51%) but not at 5 years (22% vs 22%) (overall HR 0.92, P<0.05). When patients who received systemic therapy but did not ultimately undergo resection were included, there was also no difference in survival (HR 0.98, P=0.3).

Conclusion: Hospitals that use neoadjuvant therapy more frequently in T1-2 PanCa demonstrate lower rates of nodal and margin positivity in resected patients. When patients receiving systemic therapy without ultimately undergoing resection are included, there is no survival difference across hospitals, although these patients may be spared surgical morbidity. Institutional practices regarding neoadjuvant therapy for PanCa do not result in improved aggregate outcomes.

04.20 Effects of CD47mAb Treatment on VEGF Signaling in the Huh-7 Hepatocellular Carcinoma Cell Line

Posted on January 25, 2017

D. Chirumbole1, M. Xu1, X. Wang1, G. A. Upadhya1, Y. Lin1, W. C. Chapman1  1Washington University,Transplant Section, General Surgery,St. Louis, MO, USA

Introduction:
Hepatocellular carcinoma (HCC) is the most common primary liver malignancy, and studies suggest that CD47mAb immunotherapy could be an effective therapeutic option. CD47 ligation has also been shown to inhibit vascular endothelial growth factor (VEGF) signaling in various cell types. However, few studies have considered the effects of CD47mAb treatment on VEGF signaling and tumor progression in HCC. The purpose of this study is to investigate the effects of CD47mAb treatment on VEGF signaling in the Huh-7 HCC cell line.  

Methods:
The Huh-7 HCC cell line was treated with Control IgG, CD47mAb1 (VX1000R), or CD47mAb2 (VX1004R) at a concentration of 10 µg/mL for 24 hours. Cells were then collected for study. ELISA, qRT-PCR, Western blot, and immunofluorescent staining analyses were performed to investigate the expression of VEGFs, VEGFRs, epidermal growth factor receptor (EGFR), and downstream markers related to cell growth and proliferation at the mRNA or protein level. 

Results:
VEGFR-1 and -2 expression was significantly reduced in VX1004R-treated cells compared to control in terms of relative density to β-actin (VEGFR-1: 0.181±0.105 vs. 0.618±0.064, p=0.004; VEGFR-2: 0.127±0.071 vs. 0.259±0.038, p=0.048). Phosphorylated ERK1/2 and total p38 MAPK were also significantly reduced after VX1004R treatment (phospho-ERK1/2: 0.582±0.020 vs. 0.736±0.088, p=0.042; p38 MAPK: 0.857±0.077 vs. 1.120±0.014, p=0.004), as was the expression of proliferating cell nuclear antigen (PCNA) (0.457±0.044 vs. 0.546±0.003, p=0.047). On the other hand, there was no significant change in phosphorylated p38 MAPK expression after treatment. VX1000R-treated cells showed a trend toward decreased expression of most markers, but there were no significant differences at p<0.05. Expression of vegf-a mRNA was significantly reduced in both CD47mAb-treated groups (VX1004R: fold change=0.593±0.061, p=0.002; VX1000R: fold change=0.775±0.046, p=0.010). Expression of vegf-b mRNA was reduced in VX1004R-treated cells only (fold change=0.693±0.083, p=0.010). There was a non-significant trend toward decreased EGFR expression in both CD47mAb-treated groups. 

Conclusion:
The results of this study suggest that CD47mAb treatment may inhibit VEGF signaling, downstream kinase activation, and cell proliferation in the Huh-7 cell line and therefore could inhibit tumor progression. VX1004R appears to be a more effective treatment than VX1000R. Further study is needed to investigate the effects of CD47mAb treatment on VEGF signaling in other HCC cell lines and in an in vivo system.   
 

04.16 The Innate Immune System Suppresses the Progression of Metastatic Pancreatic Cancer

Posted on January 25, 2017

A. D. Michaels1, S. J. Adair2, M. G. Mullen1, S. Nagdas2, J. B. Persily1, N. Jha1, J. T. Parsons2, T. W. Bauer1  1University Of Virginia,Surgery,Charlottesville, VA, USA 2University Of Virginia,Microbiology, Immunology, And Cancer Biology,Charlottesville, VA, USA

Introduction:  Despite a margin-negative (R0) resection, most patients with localized pancreatic adenocarcinoma (PDAC) eventually die from metastatic disease and therefore likely harbor occult metastases at the time of resection. The ability to inhibit growth of occult metastases could greatly improve survival of patients with PDAC. Prior studies suggest a role of the innate immune system in suppressing the progression of PDAC hepatic micrometasases. To test this hypothesis, we utilized a murine model of micrometastatic PDAC with patient-derived xenografts (PDX).

Methods:  Patient-derived PDAC cells expressing luciferase were injected into the spleens of two strains of mice resulting in liver metastases, followed by splenectomy to remove the primary tumor. The mouse strains used were athymic nude mice which lack cell-mediated immunity but have functioning innate immunity and NOD scid gamma (NSG) mice which lack both cell-mediated immunity as well as functional macrophages, dendritic cells, and NK cells. Metastatic tumor burden was evaluated by serial bioluminescent imaging. When disease burden threatened life, mice were sacrificed and necropsied.  Gene expression profiling (GEP) was performed on matched hepatic and peritoneal metastases from NSG mice. 

Results: The spleens of 12 nude and 12 NSG mice were injected with PDX 449 cells. There was faster and more extensive clearance of tumor cells in nude mice. By post-injection day 42, the majority of the NSG mice had extensive abdominal disease whereas only three nude mice had mild-to-moderate tumor burden. Figure 1 displays representative bioluminescent imaging at days 2 and 42. At necropsy, gross liver metastases were present in 5/12 (42%) NSG mice and 100% had peritoneal metastases. By bioluminescent imaging through 78 days, 4/12 (33%) nude mice had peritoneal metastasis and there was no evidence of any hepatic metastasis. For both metastatic sites, the NSGs had significantly more disease (liver, p=0.04; peritoneum, p=0.001). GEP of metastases from three NSG mice demonstrated significant differences between tumors from the different sites. Hierarchical clustering by innate immunity-related genes revealed that hepatic tumors cluster separately from peritoneal tumors with the most substantial differences in the TLR, VEGFA, and MAPK signaling pathways.

Conclusion: The innate immune system significantly inhibits progression of occult metastatic PDAC cells in the liver and peritoneum. Immunotherapy approaches which augment the innate immune system could greatly improve survival in patients with PDAC.

 

04.15 Cancer-Associated Plastic IL17+Foxp3+ and ex-Th17 Foxp3+ Cells are Immunosuppressive

Posted on January 25, 2017

S. E. Berkey1, D. L. Bartlett1, N. Obermajer1  1University Of Pittsburgh,Department Of Surgery,Pittsburgh, PA, USA

Introduction: Regulatory T (Treg) cells are integral in maintaining immune homeostasis and are associated with the inflammatory immunosuppression observed in cancer. Expansion of Treg cells within the tumor is a known barrier to successful cancer immunotherapy. In previous experiments, we have demonstrated that Th17 cells are a novel source of tumor-induced forkhead box P3 (Foxp3+) cells in addition to the natural (n)Treg and induced (i)Treg cells, which develop from naïve precursors. During tumor growth, Th17 cells progressively transdifferentiate into IL17+Foxp3+ and ex-Th17 IL17negFoxp3+ T cells. Here, we aimed to characterize the immunosuppressive ability of the ex-Th17 IL17negFoxp3+ T cells.

 

Methods: Using novel Th17eYFP-Foxp3mRFP reporter mice, CD4+ cells were isolated from splenocytes. The cells were cultured under Th17-driving conditions (IL6, IL23, and TGF-β) for 3-4 days, followed by Treg– driving conditions (TGF-β) for an additional 3 days, and then sorted into Th17-Treg subsets (eYFP+IL17+Foxp3neg, eYFP+IL17+Foxp3+, eYFP+IL17negFoxp3neg, eYFP+IL17negFoxp3+, eYFPnegIL17negFoxp3neg and eYFPnegIL17negFoxp3+ (inducible (i)Treg cells)). CD4+ cells were then isolated from CD45.1 mice and stained with the proliferation dye, carboxyfluorescein succinimidyl ester (CFSE). These CD4+ cells were activated with anti-CD3e Ab in the presence of CD4neg fraction and Th17-Treg subsets (2:1 ratio of CD4+ target cells: Th17-Treg subset). The cells were incubated for 72 hours and then analyzed by flow cytometry. The percentage of proliferating cells (CFSEneg, gated on live/CD45.1+CD4+ cells) was calculated as a fraction of proliferating cells compared to control condition where CD4+ cells were activated with anti-CD3e Ab in the absence of Th17-Treg subsets.

 

Results: We demonstrate that IL17+Foxp3+ and ex-Th17 IL17negFoxp3+ T cells have greater suppressive effects on CD4+ T cell proliferation compared to Th17 cells and have similar suppressive effects as (i)Treg cells.

 

Conclusions: The conversion of Th17 into Foxp3+ T cells is an alternative route of IL17+Foxp3+ and IL17negFoxp3+ T cell development in tumors. Due to their immunosuppressive properties, transdifferentiated Treg cells represent a novel target and warrant new therapeutic approaches aimed at eliminating Foxp3+ T cells in tumor immunotherapy.

04.14 Exploring Immunologic Aspects of Regional Therapy in Melanoma

Posted on January 25, 2017

J. A. Perone1, T. Tamasa1,2, M. Tsutsui2, P. Dolber1, I. V. Pinchuk1, D. S. Tyler1,2, K. Olino1  2Durham Veterans Affairs Medical Center,Durham, NC, USA 1University Of Texas Medical Branch,Surgery,Galveston, TX, USA

Introduction:

In- transit melanoma is an aggressive, rare disease subtype presenting with multiple subcutaneous lesions. Treatment options include resection, intralesional therapy, and regional chemotherapy with isolated limb infusion (ILI).  Despite short-term efficacy of ILI, disease recurrence and development of metastatic disease are common.  New research has focused on the ability of chemotherapy, particularly doxorubicin to elicit immunogenic cell death. We hypothesized that both intra-lesional (ITI) and regional treatment with doxorubicin would lead to not only treatment responses due to its cytolytic activity, but would also result in immunogenic cell death and the activation of innate immune responses. 

 

Methods:

B16 F10.9-OVA melanoma was injected intra-dermally into C57Bl/6 female mice to create a solitary leg tumor. Mice were randomized into 4 treatment groups: saline ITI, saline ILI and Doxorubicin ITI or ILI.  Mice were followed for tumor growth, survival and local toxicities. Immunologic evaluation was performed by flow cytometry and a 34-cytokine bioplex assay of the treated tumors.

 

Results:

Mice treated with doxorubicin both with ILI or ITI had delayed tumor growth (p<0.0001, one-way ANOVA) and longer survival (p<0.0001, log-rank test) when compared to saline controls. Bioplex analysis 7 days following treatment with chemotherapy showed an increase in G-CSF production within the doxorubicin treated tumors while VEGF production decreased (Figure 1).

 

Conclusion:

ITI and ILI with doxorubicin leads to a local immune response as well as direct cytotoxic effects in an orthotopic model of murine melanoma.  This suggests that an innate immune response is present likely due to both infiltration of neutrophils and macrophages.  The decrease in VEGF production is also important as this may also improve the ability within the local tumor microenvironment to overcome immunosuppressive elements.

04.13 The prognostic Impact of HLA Class I (A, B, C), HLA-E and HLA-G expression in Pancreatic Cancer

Posted on January 25, 2017

S. W. De Geus1, H. A. Prevoo1, N. G. Dekker-Ensink1, B. A. Bonsing1, H. Morreau2, A. L. Vahrmeijer1, P. J. Van Den Elsen3, C. J. Van De Velde1, P. J. Kuppen1  1Leiden University Medical Center,Surgery,Leiden, 2333 ZA, Netherlands 2Leiden University Medical Center,Pathology,Leiden, 2333 ZA, Netherlands 3Leiden University Medical Center,Immunohematology,Leiden, 2333 ZA, Netherlands

Introduction:  Pancreatic cancer has a notoriously dismal prognosis, which is to a large degree caused by its high metastatic potential. Even after complete surgical resection of the primary tumor, distal recurrence rates range from 46% to 89%. Tumor evasion and suppression of the host immune surveillance system has demonstrated to affect clinical outcome in various malignancies. However, the role of immune system in pancreatic cancer remains mainly unclear. The purpose of this study was to investigate the clinical impact of classical human leukocyte antigen (HLA) class I (A, B, C), and non-classical HLA-G and HLA-E tumor expression in pancreatic cancer patients.

Methods:  Immunohistochemical stainings were performed for HLA class I (A, B, C), HLA-G and HLA-E tumor expression in pancreatic adenocarcinoma patients (n = 137) who underwent primary surgery between 2003 and 2013. Survival analysis was performed using the multivariate Cox proportional hazard models for overall (OS) and relapse-free (RFS) survival. 

Results: HLA class I, HLA-G and HLA-E expression was observed in respectively 50%, 17% and 96% of the pancreatic adenocarcinoma patients. On multivariate analysis, primary tumor stage pT2-4 (HR, 2.188; 95% CI, 1.029-4.653; p=0.042), lymph node stage pN1 (HR, 1.327; 95% CI, 0.809-2.176; p=0.262), and HLA-G expression (HR, 1.743; 95% CI, 1.010-3.009; p=0.046) were associated with unfavorable OS. HLA-G expression resulted in median OS of 10.0 months (95% CI, 3.3 – 16.7 months), whereas absence of HLA-G expression resulted in median OS of 18.0 months (95% CI, 13.9 – 22.1 months). In addition, primary tumor stage pT2-4 (HR, 1.745; 95% CI, 0.851-3.577; p=0.128), lymph node stage pN1 (HR, 1.281; 95% CI, 0.785-2.090; p=0.321), and HLA-G expression (HR, 1.820; 95% Cl, 1.053-3.146; p=0.032) were also independently predictive for adverse RFS. Median RFS was 7.0 months (95% CI, 3.0 – 11.0 months) for patients demonstrating positive HLA-G expression compared to 12.0 months (95% CI, 8.8 – 15.3 months) for patients with absent HLA-G expression. HLA class I and HLA-E expression showed no correlated with OS or RFS in pancreatic adenocarcinoma patients.    

Conclusion: This study demonstrated that HLA-G expression was significantly associated with adverse OS and RFS in pancreatic adenocarcinoma patients. Consequently, immunohistochemistry for these markers may aid identification of patients with a poor prognosis, who likely benefit less from an upfront surgical strategy. In addition, these results provide further evidence for the immunogenic character of pancreatic cancer and subsequent potential for therapeutic strategies targeting the immune system.

 

04.10 Stat3 inhibition enhances myeloid derived suppressor cell apoptosis via Fas/FasL signaling

Posted on January 25, 2017

P. Guha1, J. C. Gardell1, M. Cunetta1, M. Lima1, J. Darpolor1, S. Ferree1,2, N. Espat1,2, S. C. Katz1,2  1Roger Williams Medical Center,Surgical Oncology,Providence, RI, USA 2Boston University,Boston, MA, USA

Introduction: Myeloid derived suppressor cells (MDSC) mediate intrahepatic immune suppression in liver metastasis (LM) patients. MDSCs are a heterogeneous group of immature myeloid cells that limit the efficacy of genetically engineered chimeric antigen receptor T cells (CAR-T).  We seek mechanistic targets to limit MDSC expansion or induce death to enhance CAR-T efficacy for LM.

Methods: Tumor burden of murine LM model treated with DMSO or STATTIC or BBI was measured using bioluminescence. Flow cytometry, western blot, qPCR and immunofluorescence was performed on MDSC isolated from tumors. LDH assay and CFSE was used to measure MDSC induced suppression of CAR-T.

Results:Previously, we have shown that the GM-CSF/Jak2/STAT3 axis drives liver MDSC (L-MDSC) expansion and suppressive function in murine LM model. We hypothesized that blocking STAT3 activation would cause decreased L-MDSC viability in the tumor microenvironment. STATTIC and BBI target STAT3 by inhibiting its activation, dimerization and nuclear translocation. Both STATTIC and BBI caused a reduction in tumor burden in our LM model with decrease in L-MDSC and pSTAT3 levels as compared to the vehicle (DMSO) control mice. When L-MDSC isolated from STATTIC or BBI treated tumor bearing mice were co-cultured with CAR-T cells, there was a significant decrease in immunosuppression as evidenced by enhanced tumor cell cytotoxicity (29+4% DMSO,54+4% STATTIC p=0.02, 52+7% BBI p=0.01) and CAR-T expansion (22+2% DMSO,36+3% STATTIC p=0.001, 38+4% BBI p=0.001). There was significant increase in apoptosis of L-MDSCs (table) isolated from STATTIC or BBI treated tumor bearing mice as compared to DMSO control. Investigation of apoptosis pathway signaling molecules showed that there was an up-regulation of pro-apoptotic Bax (STATTIC-3.6 fold and BBI-2.7 fold over DMSO, p<0.05) and down-regulation of anti-apoptotic Bcl2 (STATTIC-0.8 fold and BBI-0.5 fold over DMSO, p<0.05) protein in L-MDSCs. This was also associated with significant decrease in proliferation markers such as phosphorylated Erk (pErk) (STATTIC-0.3 fold and BBI-0.5 fold over DMSO, p<0.05), phosphorylated Akt (pAkt) (STATTIC-0.3 fold and BBI-0.4 fold over DMSO, p<0.05). Interestingly, there was an increase in phosphorylated p38 (pp38) MAPK activity (STATTIC-2.7 fold and BBI-3.2 fold over DMSO, p<0.05), Fas (table) and TNFα (STATTIC-3.4 fold and BBI-2.8 fold over DMSO, p<0.05) expression in MDSCs. 

Conclusion: Our findings indicate that STAT3 inhibition caused an increase in L-MDSC apoptosis through up-regulation of Fas/Fas-L with down-stream pro-apoptotic signaling via p38MAPK. Thus, blockade of STAT3 in tumor bearing mice may enable us to rescue the anti-tumor immunity for LM through induction of L-MDSC death.

 

04.09 Pancreatic Cancer Resistance to MEK Inhibition is Mediated by IL-1α-Dependent STAT3 Reactivation

Posted on January 25, 2017

A. A. Gaidarski1, C. Roberts1, M. VanSaun1, N. Nagathihalli1, J. Castellanos2, N. Merchant1  1University Of Miami,Department Of Surgery,Miami, FL, USA 2Vanderbilt University Medical Center,Department Of Surgery,Nashville, TN, USA

Introduction: Chemotherapy against Pancreatic Ductal Adenocarcinoma (PDAC) is daunted by innate and inducible drug resistance. Attempts to target downstream effectors of KRAS, the hallmark oncogenic mutation in PDAC, are plagued by activation of resistance pathways. We previously identified Signal Transducer and Activator of Transcription 3 (STAT3) as a mediator of resistance to MEK inhibition. However, the mechanism of STAT3 activation in response to MEK inhibition is not known. We observed decreased Interleukin 1 (IL-1) levels in response to MEK inhibition. IL-1 is a pro-inflammatory cytokine that has been broadly implicated in carcinogenesis. The full scope of IL-1 activity and its role in tumor development is not understood due the complex, pleiotropic, and often opposing effects of the distinct subtypes IL-1α  and IL-1β . In PDAC, IL-1α  stimulates both the MAPK and the NF-kB pathways, key drivers of proliferation, invasion, and drug resistance. We now demonstrate that IL-1α  down-regulates STAT3 activity. Loss of this regulation could account for development of resistance in response to MEK inhibition.

 

Methods: A cytokine array of pancreatic tumors from Ptf1acre/+;LSL-KrasG12D;Tgfbr2fl/fl mice identified reduced levels of IL-1α  and IL-1β  in response to MEK inhibition. In order to investigate whether MEK inhibition effected IL-1 signaling in pancreatic tumor cells, cultures of Panc1, a human pancreatic cancer cell line, were stimulated with recombinant IL-1α  or IL-1β  alone and treated with AZD6244, an inhibitor of MEK. Western blot analysis was performed for protein mediators of tumor proliferation and drug resistance, including phospho-NF-kB, phospho-ERK1/2, and phospho-STAT3. Gene expression of NF-kB and MAPK transcriptional targets including IL-1α , IL-1β,  and IL-6, were assayed using quantitative real-time PCR.

 

Results: Stimulation of human pancreatic cancer cell line Panc1 with IL-1α , but not IL-1β , induced phosphorylation of NF-kB and ERK1/2, consistent with established signal transduction pathways. ERK1/2 activation was effectively reduced in the presence of the MEK inhibitor. IL-1α  reduced phosphorylation of STAT3 at the classical activating tyrosine (Y705), while concurrently increasing phosphorylation of a non-classical serine (S757) moiety, which has been implicated in gating the duration of active pSTAT3-Y705. IL-1α  stimulation induced gene expression of IL-1α , IL-1β  and IL-6, which was reduce by MEK inhibition.

 

Conclusions: Our results confirm that IL-1α  and not IL-1β  signaling in pancreatic tumor cells activates NF-kB and MAPK. Additionally, we identify regulatory S757 phosphorylation of STAT3 as a novel downstream target of IL-1α . Hence, we propose that development of resistance to MEK inhibition is mediated by a loss of IL-1α -mediated regulation of STAT3, permitting reactivation of target genes and tumor growth.

04.08 The Immunomodulatory Effect of Vesicular Stomatitis Virus on Peritoneal Surface Malignancy.

Posted on January 25, 2017

G. Davis2, S. A. Northrup1, M. Westcott2, J. H. Stewart1,2  1Duke University Medical Center,Surgical Oncology,Durham, NC, USA 2Wake Forest University School Of Medicine,Winston-Salem, NC, USA

Introduction:  Oncolytic virotherapy (OV) using vesicular stomatitis virus (VSV) has become an attractive therapeutic strategy for treating peritoneal surface dissemination of colorectal cancer (PSD of CRC) because it selectively induces apoptosis in malignant cells while sparing non-malignant cells. OV’s also modulate the immune environment to induce a systemic anti-tumor response. The goal of this project was to analyze the immunomodulatory and oncolytic functions of VSV using an established murine model of PSD from CRC. We used an 8-marker flow cytometry panel to characterize the cell types and cytokine environment of the peritoneal cavity in VSV treated and control mice.  

Methods:  This study utilized an established syngeneic murine model with CT26 cell line coexpressing luciferase and GFP. In this model, 5-6 week old balb/c mice are injected intraperitonealy with 1×106 luciferase/GFP-expressing CT26 cells. On day 4, the mice were imaged utilizing an IVIS system to determine the establishment of tumor. On day 5, mice were injected with either PBS or 1×108 PFU M51R-VSV.  Mice were sacrificed at 1, 3, or 7 days after virus/mock treatment, and a peritoneal lavage was performed on each mouse to collect peritoneal exudate cells (PEC) from the cavity. Cells from each sample were processed for flow cytometry, and supernatants from each sample were collected for a multiplex cytokine bead analysis. PECs were immunotyped using an 8-color panel, supernatants were assessed using a Biolegend 13-plex Mouse Inflammation Panel, and all samples were analyzed with a BD Facs Canto. 

Results: The results from this study show VSV-mediated alterations in the peritoneal cavity as early 24 hours post virus treatment. At this early time point, M51R-VSV treated animals have significantly fewer CD4+ T cells, CD8a+ T cells, and CD19+ CD11b+ B-1 B Cells. Additionally, M51R treatment altered the cytokine microenvironment; M51R-treated cavities contained less MCP-1 than mock-treated animals. On day 7, M51R-treated mice contained approximately 75% more CD4+ T cells than their PBS-treated cohorts. There were also significant differences in CD11b+ myeloid lineages. These changes in cellular composition were accompanied by marked changes in the local concentration of IL-6 in mock-treated vs M51R-treated animals. 

Conclusion: The results from this study suggest that VSV modulates the immune response through a number of key players and facilitates its therapeutic effect by altering the innate response at early time points, in order to produce an anti-tumor adaptive response at later time points. In our model, alterations in early recruitment of T-cells and myeloid cells, along with cytolysis of tumor cells resulted in a robust CD4 helper T cell response and a marked reduction in pro-tumor myeloid lineages 7 days after virus treatment.
 

04.06 Topical Imiquimod for Cutaneous In-Transit Metastases of Melanoma

Posted on January 25, 2017

K. M. Leick1, E. P. Salerno1, J. M. Obeid1, D. H. Deacon1, C. L. Slingluff1  1University Of Virginia,Surgery,Charlottesville, VA, USA

Introduction: Topical imiquimod, a TLR7 agonist, has been reported to induce regression of some cutaneous in-transit metastases of melanoma, yet rates of tumor control and predictors of response are not established. We have reported that melanoma metastases may be grouped into Immunotypes representing patterns of immune cell infiltration: A (none), B (perivascular cuffing), and C (diffuse). The goals of this study were to provide pilot data toward testing the following hypotheses: 1) imiquimod may induce complete regressions of cutaneous melanoma metastases, 2) imiquimod extends time to major surgery, and 3) clinical responses to imiquimod are more frequent when the metastases have pre-existing immune cell infiltrates.

Methods: The UVA melanoma database was searched for patients with in-transit metastases treated with imiquimod (n=12) and matched for control patients with in-transit metastases who were not treated with imiquimod (n=11). Biopsies of metastases (IRB #10598 and 10803) were evaluable for 9 treated patients prior to initiating imiquimod. Immunohistochemistry staining (IHC) was performed on formalin-fixed paraffin-embedded sections of those tumors, using antibodies to CD45 and CD34 to assess patterns of immune cell infiltrates (Immunotypes).

Results:Among 12 patients treated with topical imiquimod, 2 (17%) had complete regressions of the metastases, while 6 (50%) had mixed responses. Complete or mixed responses typically enabled patients to delay or to avoid major surgical resections or limb perfusion. Complete responses were all in the tumors with Immunotypes B and C (2/7), and none were in those with Immunotype A (0/2). Median time to surgery for the imiquimod group (357 days) trended longer than for controls (49 days, p=0.076). For the evaluable imiquimod-treated patients, higher Immunotype (B and C vs. A) was associated with improved survival (p=0.0019).

Conclusion:These pilot data demonstrate that a small but meaningful subset of patients experience complete regressions of skin metastases with imiquimod alone. Further, the findings suggest that imiquimod therapy may provide sufficient control to avoid or to delay major surgical resections. Both complete responses were in patients whose tumors had significant immune cell infiltrates prior to initiating therapy. Overall, metastases with higher Immunotype (B or C) who were treated with imiquimod had enhanced survival. These pilot data support use of imiquimod for patients with superficial metastatic disease who are suboptimal candidates for major surgical resection or aggressive systemic therapy. These data also suggest that T cell infiltration in the metastases may be associated with better outcomes for patients treated with imiquimod.

 

04.05 Radiation and Natural Killer Cell Therapy in Canine Sarcomas: Initial Results of a Co-Clinical Trial

Posted on January 25, 2017

J. S. Park1, Z. Wang2, S. K. Grossenbacher2, J. I. Luna2, I. Sturgill2, S. S. Withers3, M. S. Kent3, M. Chen4, W. T. Culp3, R. Rebhun3, A. M. Monjazeb5, W. J. Murphy2, R. J. Canter1  1UC Davis,Surgical Oncology/Surgery,Sacramento, CA, USA 2UC Davis,Dermatology,Sacramento, CA, USA 3UC Davis School Of Veterinary Medicine,Surgical And Radiological Sciences,Davis, CA, USA 4UC Davis,Pathology And Laboratory Medicine,Sacramento, CA, USA 5UC Davis,Radiation Oncology,Sacramento, CA, USA

Introduction: We have previously shown that radiotherapy (RT) increases natural killer (NK) cytotoxicity and homing in pre-clinical models of human solid cancers, including sarcomas. Since canine clinical trials are a valuable resource for novel immunotherapy protocols and sarcomas commonly afflict dogs, we hypothesized that dog PBMC-derived NK cells would be effective in canine models of sarcoma, including adoptive transfer in a canine RT/NK clinical trial.

Methods: Canine NK cells were isolated from 15 mls of fresh whole blood using Ficoll separation and CD5 depletion. Isolated NK cells were then expanded with irradiated K562c9IL21 for 2-3 weeks. Using 6-month metastasis-free survival as the primary endpoint, we are evaluating RT and adoptive NK immunotherapy in a canine clinical trial. For this trial, treatment consists of palliative RT weekly x4 followed by two intra-lesional injections of autologous NK cells. In correlative studies, including dog patient-derived xenografts (PDX), we assessed NK homing using eFluor 670 cell proliferation dye and NK function by expression of activation markers IFNγ, granzyme B, and perforin.

Results: We have treated 8 of planned 14 dogs with osteosarcoma on protocol with a median of 76×10^6 cells (92% viable). Of 3 evaluable dogs who have reached the 6-month primary endpoint, we have observed 1 partial response and 2 are metastasis-free, including 1 dog with complete resolution of a suspicious 3 mm pulmonary nodule. In dog patients on trial, phenotyping of expanded NK cells from all patients showed > 90% granzyme B and IFNγ expression prior to adoptive transfer. Tagging experiments 1 week after intratumoral injection revealed that 11 – 60% of CD45+ cells are eFluor 670 positive, confirming persistence of injected NK cells post injection. Analysis of unactivated circulating PBMCs post-injection demonstrated a significant increase in granzyme B expression (2.25X ± 0.42, P<0.01). Dog PDX studies demonstrate that focal RT increases NK homing to sarcomas on average 3.8X±0.3 (P<0.001) compared to unirradiated controls. Immunohistochemical analysis of tissue samples post RT shows a significant increase in CD3+ tumor-infiltrating lymphocytes post RT (P<0.05, see figure). Co-culture experiments of dog PDX sarcomas ex vivo with allogeneic NK cells shows RT-induced sensitization to NK killing at doses of 10 – 20 Gy (P<0.01)

Conclusion: RT and NK immunotherapy appear synergistic in dog models of sarcoma. Preliminary results from a canine clinical trial of palliative RT and autologous NK transfer for osteosarcoma are promising, including possible abscopal effects. Further evaluation of this novel radio-immunotherapy approach is warranted.

 

04.04 Myeloid Derived Suppressor Cells in the Inflammatory Response of Pancreatitis and Pancreatic Cancer

Posted on January 25, 2017

N. E. Cieza Rubio1, R. L. Heimark1,2  1University Of Arizona,Department Of Surgery,Tucson, AZ, USA 2The Arizona Cancer Center,Tucson, AZ, USA

Introduction:

Tumor-infiltrating myeloid-derived suppressor cells (MDSCs), are important mediators of a tumor-permissive microenvironment that contributes to tumor growth and could account for the limited success of immunotherapeutic strategies. MDSCs suppress adaptive immunity by blocking T cell activation, inducing Treg accumulation, and inhibiting natural killer cell cytotoxicity against tumor cells. Aim: We investigated the roles of MDSCs in the regeneration of the exocrine pancreas associated with acute pancreatitis and the progression of acinar to ductal metaplasia. 

Methods:  

Model of Acute Pancreatitis: Pancreatitis was induced using a regimen of 7 hourly intraperitoneal injections of caerulein (50 g/kg) followed by a 48-hour rest period and a second regimen of 7 hourly intraperitoneal injections of caerulein before sacrifice at different time points (1, 12, 24, 48 and 72 hours after last caerulein injection). 

Model of MDSCs Depletion: Wild type mice were treated with bi-daily intraperitoneal doses of CXCR2a (SB-265610) (4 mg/kg). Treatment started one day before induction of pancreatitis with caerulein and continued until animal euthanasia.

Positive Selection of gMDSCs: After preparation of a single-cell suspension from pancreas or spleen, CD11b+Gr1+ cells were indirectly magnetically labeled with Anti-Ly-6G-Biotin and Anti-Biotin Microbeads.The magnetically labeled Ly6G+ cells, also known as positively selected cell fraction, were retained within the column and collected once the column is removed from the magnetic field.

Cell surface marker analysis was performed by flow cytometry using the BD FACSCANTO II.

Results:

Acute pancreatitis was induced in wild type and P48+/Cre;LSL-KRASG12D mice using caerulein and an early influx of MDSCs into the pancreas was observed flow cytometry and immunocytochemistry. Numbers of Gr1(+)CD11b(+) MDSCs increased over 20-fold in pancreata of mice with acute pancreatitis to account for nearly 15% of intrapancreatic leukocytes and have T cell suppressive properties. This marked accumulation of MDSCs returned to normal values within 24 hours of the insult in wild type mice; however, in the oncogenic KRAS mice, MDSCs levels remained elevated. When intrapancreatic MDSCs were depleted by administration of a CCR2 antagonist (SB265610) in wild type mice the severity of acinar damage was increased. This was also accompanied by a delayed regeneration determined morphologically and with the mitotic immunomarker phospho-histone H3. Isolated intrapancreatic MDSCs from treated mice induce naïve acinar cells to undergo acinar ductal metaplasia when co-cultured in collagen 3D cultures. Purified splenic MDSCs failed to induce the phenotypic transdifferentiation.

Conclusion:

MDSCs are required for adequate pancreatic regeneration in wild type mice with acute pancreatitis and their persistent elevation in oncogenic KRAS mice is not only associated with immune-evasion, but may also function as direct enhancer of malignant proliferation.

04.03 Loss of Class I MHC Adds Prognostic Value to PD-L1 Expression in Non-Small Cell Lung Cancer Patients

Posted on January 25, 2017

J. M. Obeid1, G. Erdag4, R. D. Gentzler2, M. G. Brown5, J. V. Cross3, D. H. Deacon1, T. N. Bullock3, C. L. Slingluff1  1University Of Virginia,Surgery Department,Charlottesville, VA, USA 2University Of Virginia,Heme/Onc / Medicine Department,Charlottesville, VA, USA 3University Of Virginia,Pathology Department,Charlottesville, VA, USA 4Johns Hopkins University School Of Medicine,Pathology Department,Baltimore, MD, USA 5University Of Virginia,Immunology,Charlottesville, VA, USA

Introduction:  Expression of PD-L1 on tumor cells reflects the presence of an active immune response, and its increased expression is associated with improved patient survival in several cancers. In Non-small cell lung cancer (NSCLC), however, this association is less clear. Effective immune responses may be inhibited in the tumor microenvironment by downregulating class I MHC (MHC-I) expression on cancer cells. Thus, we hypothesized that low MHC-I expression may add additional negative prognostic value in NSCLC with low PD-L1 expression.

Methods:  A tissue microarray (TMA) was constructed from 151 NSCLC specimens (90 adenocarcinoma (AdCA), 58 squamous cell (SQCC), 3 mixed histology). Patients had stage I (61%), II (22%), III (14%) and IV (3%) disease that was resected between 2011 and 2014 (median follow up: 27 months). The TMA was analyzed by immunohistochemistry for PD-L1 (clone 5H1) and MHC-I heavy chains (clone HC-10). PD-L1 expression was assessed as a percentage of cancer cells staining positive, with 5% and 50% thresholds for positivity. Expression of the MHC-I was reported as a score of 1 to 5, accounting for staining intensity and percent of tumor cells staining. Kaplan Meier curves were used to plot survival, and a log rank test was used to test associations with patient survival. 

Results: Low PD-L1 expression, below a 5% threshold, was associated with worse overall survival (p=0.02) but the association was not significant with a 50% threshold (p=0.06). Low MHC-I expression on tumor cells (score ≤ 3) was not significantly associated with worse survival (p=0.14). However, when combined, the tumors with both low MHC-I and low PD-L1 expression (<5% and <50%) were significantly associated with worse patient outcomes (p=0.002 (Figure) and p=0.004 respectively), similar results were obtained in SQCC and AdCA subsets separately, and when controlling for stage.

Conclusion: Co-occurrence of low PD-L1 expression and MHC-I downregulation in tumor cells, is associated with a significant decrease in patient survival for patients with NSCLC. When MHC-I expression is high, PD-L1 expression loses its prognostic significance. This finding may help to clarify the role of pre-existing immunity to NSCLC, and may help to develop a useful biomarker combination. By characterizing the immune profile of NSCLC, modifications of this analysis should be explored for their ability to predict patient response to immune therapy.

 

« Previous Page
Next Page »