Hiromichi Miyagaki

Overexpression of PFTK1 predicts resistance to chemotherapy in patients with oesophageal squamous cell carcinoma

Background:Recently, PFTK1 was identified as a member of the cyclin-dependent kinase family; however, its expression and clinical significance in oesophageal squamous cell carcinoma (ESCC) have not been evaluated.Methods:PFTK1 expression was initially examined by expression microarray in 77 ESCC patients. Using independent samples of 223 patients, PFTK1 expression was evaluated immunohistochemically to assess the relationship between expression and various clinicopathological parameters. The association between PFTK1 and the response to chemotherapy was also investigated in pretreatment samples of 85 patients who received chemotherapy as first treatment.Results:Significant upregulation of PFTK1 expression was noted in ESCC compared with normal epithelium. PFTK1 expression was positive in 51.6% (115 out of 223) of the tumours, but did not correlate with any clinicopathological parameter. The 5-year overall survival rate was poorer in patients positive for PFTK1 (43.6%) than those with negative expression (66.2%, P<0.001). Uni- and multivariate analyses identified PFTK1 as an independent marker of prognosis (RR=2.428, 95% CI=1.615–3.711, P<0.001). Out of 85 biopsy samples, 40 (47.1%) tumours showed PFTK1-positive expression, and the response rate to chemotherapy was significantly lower than PFTK1-negative tumours (27.9% vs 72.1%, P<0.001).Conclusion:PFTK1 is not only useful as a prognostic marker, but also as a predictor of the response to chemotherapy.

Plasma chitinase 3-like 1 is persistently elevated during first month after minimally invasive colorectal cancer resection

AIM To assess blood chitinase 3-like 1 (CHi3L1) levels for 2 mo after minimally invasive colorectal resection (MICR) for colorectal cancer (CRC). METHODS CRC patients in an Institutional Review Board approved data/plasma bank who underwent elective MICR for whom preoperative (PreOp), early postoperative (PostOp), and 1 or more late PostOp samples [postoperative day (POD) 7-27] available were included. Plasma CHi3L1 levels (ng/mL) were determined in duplicate by enzyme linked immunosorbent assay. RESULTS PreOp and PostOp plasma sample were available for 80 MICR cancer patients for the study. The median PreOp CHi3L1 level was 56.8 CI: 41.9-78.6 ng/mL (n = 80). Significantly elevated (P < 0.001) median plasma levels (ng/mL) over PreOp levels were detected on POD1 (667.7 CI: 495.7, 771.7; n = 79), POD 3 (132.6 CI: 95.5, 173.7; n = 76), POD7-13 (96.4 CI: 67.7, 136.9; n = 62), POD14-20 (101.4 CI: 80.7, 287.4; n = 22), and POD 21-27 (98.1 CI: 66.8, 137.4; n = 20, P = 0.001). No significant difference in plasma levels were noted on POD27-41. CONCLUSION Plasma CHi3L1 levels were significantly elevated for one month after MICR. Persistently elevated plasma CHi3L1 may support the growth of residual tumor and metastasis.

Abstract 4925: Plasma levels of cartilage oligomeric matrix protein (COMP), are significantly increased in patients with colorectal cancer

Introduction: Cartilage Oligomeric Matrix Protein (COMP) is a member of the thrombospondin family of extracellular calcium-binding proteins that plays role in the assembly of extracellular matrix. COMP maintains the structural integrity of cartilage through its interaction with a number of extracellular matrix proteins. COMP has a unique binding site for Vitamin D, indicating that it may also participate in storage and delivery of cell-signaling molecules. An expression array analysis of colorectal cancer samples vs normal tissue has noted overexpression of COMP mRNA. This study was undertaken to assess preoperative blood levels of COMP in the setting of colorectal cancer(CRC). Our hypothesis is that plasma levels of COMP may be elevated due to tumor overexpression and that this protein may have potential as a diagnostic marker for colorectal cancer. Method: Patients (pts) undergoing elective colorectal resection for CRC or benign colonic pathology (BCP) that had been prospectively enrolled in an IRB approved tissue and data bank for whom adequate PreOp plasma samples were available were studied. Demographic, clinical, and pathologic data were collected. Plasma COMP levels were determined via ELISA in duplicate and are reported as median +95%CI (ng/ml). Tissue expression levels were determined in paired tumor and normal tissue samples of a subpopulation of study pts by QRT-PCR .The candidacy of COMP as a diagnostic marker for CRC was validated by the receiver operating characteristic (ROC) curve and by the area under the ROC curve (AUC) results. The Wilcoxon/Kruskal-Wallis test was used for statistical analysis, (significance p Results: A total of 151 CRC (73%colon, 27%rectal) and 73 BCP pats (Adenoma 21%, diverticulitis 57%, other 22%) were studied. The male/female ratios were similar. The CRC stage distribution was: Stage 1, 22%; Stage 2, 32%; Stage 3, 31%; and Stage 4, 15%. The median plasma COMP levels were significantly higher in the CRC pts (209.2,CI: 187.8,234) vs. the BCP pats(129.6,CI: 110.7,149.2; P = Conclusion: The CRC median plasma COMP level was 61% higher than the median BCP result and the Stage 3 median level 50% higher than the median Stage I level. In general, plasma COMP levels increased with advancing cancer stage. The AUC results suggest COMP may have value as a CRC prognostic marker, perhaps, in combination with other protein markers to increase the sensitivity. Tissue expression analysis suggests that the most likely source of the added COMP in the plasma is the tumor itself. Further study with larger populations of control and CRC pts is warranted. Citation Format: Hmc Shantha Kumara, Hiromichi Miyagaki, Xiaohong Yan, Linda Njoh, Vesna Cekic, Nipa D. Gandhi, Melissa M. Alvarez-Downing, Richard L. Whelan. Plasma levels of cartilage oligomeric matrix protein (COMP), are significantly increased in patients with colorectal cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4925.

Mo1887 Three Protein Combinations Angiogenin, Galectin-3 and Activin-A Improves Sensitivity and Specificity of Plasma Analysis for Colorectal Cancer Detection

Introduction: Colorectal cancers (CRC) express Angiogenin (ANG), Galectin-3 (Gal-3) and Activin A (Act-A). The binding of ANG to actin on endothelial cell promote EC migration and angiogenesis. Cellular Gal-3 levels associated with cancer cell invasion, angiogenesis and tumor progression. Act-A act via binding to trans membrane receptors and support cancer cell migration. Blood levels of ANG, Gal-3 and Act-A in colorectal cancer (CRC) have not been well studied. This studys purpose was to measure preoperative (PreOp) plasma ANG, Gal-3 and Act-A levels in CRC and benign pathology (BP) patients (pts.) and to assess the diagnostic efficacy of these proteins alone and together. Method: CRC or BP pts having bowel resection for whom PreOp plasma was available (from IRB approved tissue bank) were studied. Plasma ANG (ng/ml), Gal-3 and Act-A (pg/ml) levels were analyzed in duplicate via ELISA (results: median + 95%CI). Intergroup levels were compared by the Mann-Whitney test (significant;p <0.05). The plasma receiver operating characteristic (ROC) curve and area under the ROC curve (AUC) were used to evaluate single and combined proteins levels. Results: Plasma from 43 BP (polyp 28%, diverticulitis 56%, other 16%) and 120 CRC (83% colon, 17% rectal) pts. were studied. The CRC stage distribution was: Stage1, 25%; Stage-2, 37%, Stage-3, 26%, Stage4, 12%. Median PreOp proteins levels in CRC pts. were significantly higher than BP levels.[ ANG; 390.9,CI: 373.2,405.5 vs 316.8,CI: 297.5,350.9; Gal-3; 12.0,CI: 10.6,13.8 vs 8.9,CI: 7.2,11.1; Act-A; 369.7 CI:349.2,424.0 vs 277.9 CI: 209.3,346.7;P<0.001). Plasma Act-A levels were significantly higher in the Stage 4 pts. than in the stage 1 group ( p=0.001). The single AUC values from the ROC curve for ANG, Gal-3 and Act-A were 0.719, 0.709 and 0.741 with associated 81%, 47% and 42%. The 3 protein combination improved the AUC (0.856) and specificity (90.7%). Conclusion: CRC median ANG, Gal-3 and Act-A levels were significantly higher (23%, 35% & 33% respectively) than BP levels. Although not proven, we believe the plasma elevations are due to the tumor. Higher levels of ANG, Gal-3 and Act-A in plasma of CRC pts. may be related to the tumor cells, stromal cells, and inflammatory cells surrounding the cancer; these elevated levels may related to neovascularization and inflammation-induced tissue remodeling at tumor sites. The 3 protein combination had improved AUC & specificity vs single protein results and may have value as a diagnostic panel. A larger study is needed.

Occurrence and timing of organ space surgical site infections (SSI) in rectal cancer patients: A NSQIP database review.

552 Background: This study’s aim was to determine the incidence and date of presentation of organ space SSI (in this setting most likely due to anastomotic leak or pelvic abscess) in rectal cancer resection patients. A second purpose was to determine which variable(s) impacted the timing of the presentation. Methods: The NSQIP database was queried from 2007 to 2011 for elective rectal cancer resections. Exclusion criteria included; ASA 5 status, preoperative(preop) shock/sepsis, SIRS, and recent surgery. Demographic parameters, comorbidities, lab data, and incidence and timing of organ space SSI were assessed. The statistical methods used were Fisher’s exact test (categorical variables) and Wilcoxon’s rank tests (continuous variables). Results: 8,093 rectal cancer patients were identified (Male/Female; 60.1%/39.9%; median age, 60; APR, 28 %; sphincter saving procedures,72%). The incidence of organ space SSI was 6 % (485 patients; 2.7% presented before and 3.3% after discharge). The rate was significantly ...

Tu1534 Morbidity and Outcomes of Colorectal Surgery in the Underweight Population: Results From the American College of Surgeons National Surgical Quality Improvement Program (ACSNSQIP) Database

PURPOSE: Whereas, the impact of morbid obesity on colorectal resection (CR) outcomes has been studied, there is limited data concerning CR outcomes in the underweight population (BMI ,18.5). This studys goal was to assess the underweight population (UN) that comes to CR and CR-related morbidity. METHODS: The ACS NSQIP database was queried from 2005-2010 for the CR codes. Patients (pts) who, preoperatively (preop), were ventilator dependent, ASA 5, hypotensive, had SIRS, sepsis, and emergent surgery were excluded. Pts were divided into 3 BMI groups: UN, BMI,18.5; Normal/Obese (NO), BMI ≥18.5 to ≤40; and morbidly obese (MO), BMI .40. Demographic parameters were assessed as well as surgical indications, comorbidities, preop laboratory data, and complications (including surgical site infections (SSI), transfusions, reoperation, etc). The statistical methods used were two sample for population proportions for categorical variables and Wilcoxon rank sum tests for continuous variables. RESULTS: A total of 73,516 CR pts were identified; the breakdown as per BMI was UN, 2,180, 3%; NO, 67,732, 92%; MO, 3,604, 4.9%. In the UN group there is a significantly higher proportion of colitis/enteritis pts (UN 20.4%, NO 9%, MO 5.2%) and obstruction/volvulus pts (UN 8.3%, NO 3.3%, MO 2.5%) as well as a much lower proportion of diverticulitis cases (UN 8.4%, NO 21.1%, MO 23.4%). The percent of UN cancer pts (44%) was modestly but significantly lower than the NO (47%) or MO (47.6%) groups yet the UN group had more disseminated cases. More UN pts reported weight loss (21.6%) than in the NO (4.7%) or MO (1.7%) groups. Also, more UN pts (13.6%) were on steroids (vs NO, 6.1%; MO, 4.5%, p ,0.05). Finally, the UN group had a significantly lower mean albumin level and hematocrit. Significantly fewer UN CRs were done using laparoscopic (LAP) methods than in the other groups (UN 34%, NO 45%, MO 39%). There were also more total colectomies and Hartmans procedures in UN group (p,0.05). There were significantly more complications in the underweight group (UN)(20.4%) than in the NO group (15.6%), yet, there was no difference between the UN and MO groups (20.1%). The UN groups rate of transfusions, sepsis, and reoperations were higher than noted in the NO group. Of note, there were significantly fewer superficial surgical site wound infections in the UN vs the other 2 groups (UN 5.6%, NO 8.1%, MO 16%). CONCLUSIONS: The complication rate was notably higher in the UN group which may be related to the higher incidence of colitis and obstruction cases and the greater percentage of pts with weight loss and steroid use all of which are associated with high complication rates. Diverticulitis is rare in UN pts. For unclear reasons laparoscopic methods were used less often in the UN group.

Tu1755 Microarray Analysis of T-Lymphocyte Gene Expression After Colorectal Resection

Introduction: Previous studies have established that surgical trauma is associated with significant transient alterations in cell-mediated immune function. Surgery-related immunosuppression may impact the patients ability to deal with infection. Also, tumor growth has been shown in murine studies to be increased after surgical trauma. Cell-mediated immunosuppression after resection of a primary tumor may impair the hosts ability to eradicate or contain residual tumors cells. This microarray study of perioperative T-lymphocyte (TLC) gene expression was undertaken in an effort to better understand the impact of colorectal resection (CR) on cell-mediated immune function. Method: Patients who underwent elective laparoscopic right hemicolectomy (RHC) for benign colonic disease (BCD) who had enrolled in an IRB approved blood/data bank for whom frozen preand postoperative TLCs were available were eligible for this study. Benign pathology patients were chosen in order to determine the impact of surgical trauma alone, independent of the potential effects of a cancer on immune function. Preoperative (PreOp) and postoperative day1 (POD1) blood samples were utilized. TLCs were isolated from the blood using a combination of gradient centrifugation and magnetic micro-bead separation. TLCs were subsequently lysed and total RNA extracted. cRNA was made from RNA hybridized to HG-U133APLUS oligonucleotide array. PreOp vs POD1 expression data was analyzed via Limma paired analysis to find differently expressed genes. (p.0.05 significant) and consistency of significance was analyzed via Empirical Bayes statistics (B>0 Sig.). Clinical data is presented as mean ± SD. Results: Nineteen patients (12 males/7 female, mean age 65.8± 12.8 years) met the entry criteria. The mean incision length was 7.8± 3.5cm and mean length of stay was 6.3±2.6 days. All TLC expression data met the affymetrix data QC standards. A total of 39 genes showed significant changes on POD1; 21were up-regulated and 18 were down regulated (B=0.055.3). The expression changes of 7 genes in this group were strongly significant (ABCG-1, TMEM49, FAM100B and PIM1 were upregulated and IFI44L, STAT1 and UCP2 were down regulated; P =0.02 and B =4.1-5.3). Enrichment analysis confirmed that these gene changes were likely to have significant effects on 7 signaling pathways and 3 functional categories i.e.; cell proliferation, hematological function and immune response. Conclusion: Surgical trauma affected gene expression of circulating TLCs in the immediate postoperative period. Altered gene expression may impact TLC growth and proliferation as well as immune function. These changes must be validated at the protein level and additional patients studied. Also, the duration of these changes, after surgery, must be determined. Finally, a similar study in cancer patients is also needed.